TILens Daily Edition 2026-08-31 Filters: topic=ai, github=hidden Stats: 367 articles, 21 sources, 342 research papers Top topics: AI, AI Research, Python ## News - Amazon EC2 R9g and R9gd memory optimized instances are now available Source: aws-whats-new Topic: AWS (+AI) URL: https://aws.amazon.com/about-aws/whats-new/2026/08/amazon-ec2-r9g-and-r9gd-memory-optimized-instances-are-now-available/ - Improving our alignment and security efforts Source: anthropic-blog Topic: AI Labs (+AI) URL: https://www.anthropic.com/news/improving-alignment-security-efforts - The Cheapest Token You Never Send: Agentic AI Search on MongoDB + Voyage AI Source: mongodb-blog Topic: MongoDB (+AI) URL: https://www.mongodb.com/company/blog/technical/cheapest-token-you-never-send-agentic-ai-search-on-mongodb-voyage-ai - Polimill builds Japan's next-generation public AI infrastructure Source: openai-blog Topic: AI Labs (+AI) URL: https://openai.com/index/polimill - OpenAI supports California’s bill to advance youth AI safety Source: openai-blog Topic: AI Labs (+AI) URL: https://openai.com/index/supporting-california-bill-advance-ai-youth-safety - A milestone in expanding access to AI Source: openai-blog Topic: AI Labs (+AI) URL: https://openai.com/index/expanding-access-to-ai-with-chatgpt-ads - Exclusive Early Access: Final 3 Chapters of The Hundred-Page Deep Reinforcement Learning Book Source: true-positive-weekly Topic: AI (+AI Research) URL: https://aiweekly.substack.com/p/exclusive-early-access-final-3-chapters - SpaceX is in an “enviable position”: why Anthropic is sticking with Cursor as OpenAI cuts access Source: newstack Topic: DevOps (+AI) URL: https://thenewstack.io/anthropic-spacex-cursor-compute-openai/ - Anthropic Users Hit by Infostealer Attacks, Session Thefts Source: darkreading Topic: DevSec (+AI) URL: https://www.darkreading.com/cyberattacks-data-breaches/anthropic-users-infostealer-attacks-session-thefts - OpenAI wants to charge only when AI gets it right — here’s the catch Source: newstack Topic: DevOps (+AI) URL: https://thenewstack.io/openai-outcome-based-pricing/ - The Hugging Face hack could indicate cultural issues at OpenAI Source: mit-ai Topic: AI URL: https://www.technologyreview.com/2026/08/31/1143180/hugging-face-hack-could-indicate-cultural-issues-at-openai/ - AI Model Rules Are Not Security Controls Source: darkreading Topic: DevSec (+AI) URL: https://www.darkreading.com/cyber-risk/model-knowing-rules-is-not-security-control - LLM Tool Call Reasoning with Embabel AI Source: java-code-geeks Topic: Java (+AI) URL: https://www.javacodegeeks.com/llm-tool-call-reasoning-with-embabel-ai.html - OpenShot 4.0 released Source: lwn Topic: Open Source (+AI) URL: https://lwn.net/Articles/1091606/ - Hugging Face’s robot duck is already a hit Source: the-rundown-ai Topic: AI (+AI Tools) URL: https://www.therundown.ai/articles/hugging-face-robot-duck-is-already-a-hit - Import AI 471: Why Hugging Face worries me; space mining; FIve Eyes on AI Source: importai Topic: AI URL: https://importai.substack.com/p/import-ai-471-why-hugging-face-worries - Anthropic’s Claude fixed all 10 alignment failures. Then it tried to cheat 2.4% of the time. Source: newstack Topic: DevOps (+AI) URL: https://thenewstack.io/claude-automated-alignment-research/ - DeepSeek’s first vision model vs. Gemini 3.7 Flash: It comes down to spend vs. speed Source: newstack Topic: DevOps (+AI) URL: https://thenewstack.io/deepseek-gemini-vision-comparison/ - Foundry Model Router Expands from Two Regions to 28, Refreshing Its Model Pool Source: infoq-architecture Topic: Architecture (+AI) URL: https://www.infoq.com/news/2026/08/foundry-model-router-regions/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=Architecture+%26+Design - OpenAI cuts out SpaceX-owned Cursor Source: the-rundown-ai Topic: AI (+AI Tools) URL: https://www.therundown.ai/articles/openai-cuts-out-spacex-owned-cursor - OpenAI blocks Cursor 🚫, Apple's new CEO 🍎, GitHub Agentic Workflows 👨‍💻 Source: tldr-tech Topic: General (+AI) URL: https://tldr.tech/tech/2026-08-31 ## Research - TimesFM-3: A zero-shot foundation model for multivariate forecasting Source: google-research-blog Topic: AI Labs (+AI, AI Research) URL: https://research.google/blog/timesfm-3-a-zero-shot-foundation-model-for-multivariate-forecasting/ - Your LLM Can Return Perfect JSON and Still Be Wrong Source: towards-data-science Topic: AI Research (+AI, Python) URL: https://towardsdatascience.com/your-llm-can-return-perfect-json-and-still-be-wrong/ - FAQ as RAG: When You Get to Design the Corpus Source: towards-data-science Topic: AI Research (+AI, Python) URL: https://towardsdatascience.com/faq-as-rag-when-you-get-to-design-the-corpus/ - Why RAG Complexity Should Be Earned Source: towards-data-science Topic: AI Research (+AI, Python) URL: https://towardsdatascience.com/why-rag-complexity-should-be-earned/ - AgentOps Is Not MLOps: What Breaks in Your Monitoring Stack When Agents Go to Production Source: towards-data-science Topic: AI Research (+AI, Python) URL: https://towardsdatascience.com/agentops-is-not-mlops-what-breaks-in-your-monitoring-stack-when-agents-go-to-production/ - PersonaForge: Realistic Multi-Turn User Simulation for Agentic Systems Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28378 - Cognitive Chain-of-Thought (CoCoT): Structured Multimodal Reasoning about Social Situations Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2507.20409 - SimpCue: Cue-Based Prompting for Multilingual Text Simplification Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28042 - MemoryCard: Topic-Aware Multi-Modal Clue Compression for Long-Video Question Answering Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.05917 - Are These Modules Worth Their Cost? A Paradigm-Level Accuracy-Cost Analysis of In-context Learning Text-to-SQL Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28432 - Steering Multimodal Large Language Models Decoding for Context-Aware Safety Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.19212 - DisCTI: Who Needs to Know Timely? Automated Sector-Aware Cyber Threat Intelligence Dissemination Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27967 - Semantic Head Specialization Guides Hybrid ViT Attention for Multimodal LLMs Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28383 - Semantic Watermarking with Order-Robust Detection over Sub-sentence Units Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27666 - ProfileFoundry: A Synthetic Person-Object Substrate for Privacy, Memory, and Tool-Use Evaluation in LLM Agent Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.26403 - PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selection Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2502.12119 - Auditing Generative Audio Calls for Known-Task Audio-LLM Evaluation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27817 - Does Finetuning with Scientific Data Increase Hallucinations? A Multi-domain Factuality Evaluation of LLMs Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.21359 - Select, Don't Train: The Benefits of Modular Entity Disambiguation with LLM-Based Selection Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27470 - CEDAR: Automata as Verifiable Interfaces for Language-Guided Embodied Action Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27797 - Search, Inspect, Fetch: Exploiting Structure-Aware Boolean Retrieval for Deep-Search Agents Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.02751 - The Effect of Emotional Context on Large Language Models' Endorsement of Premature Decisions: Comparing Emotional Vulnerability Across Six Commercial Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27465 - Select, Label, Evaluate: Active Testing in NLP Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.21840 - Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28496 - The Company You Keep: How LLMs Respond to Dark Triad Traits Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.04299 - Long Story Short: Story-level Video Understanding from 20K Short Films Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2406.10221 - FENCE: A Financial and Multimodal Jailbreak Detection Dataset Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.18154 - CNeo-Bench: Diagnosing Large Language Models on Chinese Neologisms Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28053 - CoFrGeNet: Continued Fraction Architectures for Language Generation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2601.21766 - A Shaky Voice Is Not Always a Dodge: Benchmarking Textual and Vocal Evasion Detection in Earnings Calls Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28040 - Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layers Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28327 - PersonaEdit: Representative Sample Selection for Personalized Model Editing Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27816 - UIC-AIHealth4All at ArchEHR-QA 2026: Answer-First Evidence Grounding for Clinical Question Answering Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27467 - Securing Multi-Agent GIS Systems: Risk Evaluation and Prompt Hardening Optimization Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.17092 - First Make It Playable, Then Make It Good: Staged Interaction Learning for Small Dialogue-Game Agents Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27672 - AI Models Can Predict and Collaboratively Modulate Human Memory Search Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26152 - Synthetic Linguistic Agency: How an Embodied Mortal Agent Learns Linguistic Affordances through Consequential Social Experience Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27843 - INSPIRE: An Internalize-Then-Improve Approach for Example-Driven Mathematical Reasoning Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27501 - NL2AGBench: Benchmarking LLM Auto-Formalization for AlphaGeometry Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28481 - Beyond the Rabbit Hole: Mapping the Relational Harms of QAnon Radicalization Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2601.17658 - Compositional Failure in Audio-Visual LLMs: Late-Layer Prior Dominance Under Cross-modal Conflict Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27785 - SURE-Challenge: Evaluating Speech Evidence Before Speech-LLM Generation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27783 - Load-Bearing Context: The Question Damage Score for Evaluating Context Reliance in Linguistic Reasoning Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27756 - AI Alignment through a Game-theoretic Lens: A Survey Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27910 - LandingAgent: A Reference-Annotated Dataset and Agentic Generation Framework for Landing Pages Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27902 - CRAM: Centroid-Routing and Adaptive MoE for Multimodal Continual Instruction Tuning Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.02502 - Nested Byte-Level Vocabularies Are Cheap to Deploy and Expensive to Share: A Pre-Registered Negative Result Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28151 - Informational Antilocality and the Locality Bias in LLMs Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27760 - ElementCheck: Complexity-Aware Long-Form Text Factuality Evaluation via Sentence Elements Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26118 - CASPER in the Machine: Insights into Character Variety in LLM-Generated Stories Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.22454 - Human Label Variation as Stable Signal: Learning Annotator-Specific Explanation Behavior via Cross-Annotator Preference Optimization Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.28802 - Representing and Parsing Korean Constituency Structure at Different Levels of Granularity Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27035 - Fidelity Is Not Enough: Dispatch-Level Instrumentation for Agentic Datasheet Extraction Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28439 - Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28009 - Quantifying Affective Bias in Low-Resource Media: Large-Scale Emotion Profiling of Bengali Headlines Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2510.17252 - Large Reasoning Models Struggle to Transfer Parametric Knowledge Across Scripts Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.17070 - When Linguistic and Internal Confidence Diverge in Large Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28382 - A Unified Framework to Elicit Structured Feedback for Interpretable Multi-Trait Essay Scoring Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28407 - CamoDocs: A Poisoning Attack Against Retrieval-Augmented Language Models Using Camouflaged Documents Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28389 - Benchmarking LLM-as-a-Judge for Long-Form Output Evaluation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.01629 - PRISM: Agentic Retrieval with LLMs for Multi-Hop Question Answering Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2510.14278 - H-Scale: Hessian-Guided Scale Refinement for NVFP4 Sub-Byte LLM Inference Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28113 - Text Restoration of Ancient Documents with Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28170 - When Robots Mishear Us: Mapping the Safety Risks of Voice-Controlled Embodied AI Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28518 - The Granularity Gap: A Multi-Dimensional Cross-Generational Audit of Sycophancy in Gemini Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.05183 - Entity-Memory Graph Retrieval Improves Evidence Coverage in Long-Conversation Question Answering Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27925 - Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledge Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28478 - Do LLM Agents Mirror Socio-Cognitive Effects in Power-Asymmetric Conversations? Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.17694 - Knowing Before Answering: Decoding Language Models for Reliable RAG Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27661 - Embedding Models for Stance-Aware Argument Retrieval Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28283 - CultureConverse: A Multilingual Multi-turn Simulation Harness for Culturally Grounded Assistance in East and Southeast Asia Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28405 - Learning a Single Token to Replace Long System Prompts in LLMs Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.23271 - When Tokenizers Fail: Byte-Level Chunking for Zero-Shot Transfer to Low-Resource Languages Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27658 - The Telephone Game: Evaluating Semantic Drift in Unified Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.04438 - A Wolf in Sheep's Clothing: Targeted Routing Hijacking in Federated RAG Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.28112 - Speculative Probing: LLM Monitoring at Speculative-Decoding Cost Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28099 - Lexically conditioned realization ambiguity in Korean predicate morphology Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27966 - Stranger, Fan, or Peer? A Systematic Study on the Role of Interlocutor in Persona-Based Dialogue Generation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28467 - Comparing Chunking and Embedding Strategies for Turkish RAG Systems Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26192 - Psychologically Potent, Computationally Invisible: LLMs Generate Social-Comparison-Eliciting Posts They Fail to Detect Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.01017 - Auditing LLM Benchmarks with Item Response Theory Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.30504 - Set-shifting Behavioral Test for Harnessed Agents Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.13396 - Predicting Turn-Taking Outcomes in Multi-Party Conversation: Interpretable Modelling of Speech and Gaze Dynamics with Interpersonal Closeness Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27988 - What Makes Agent Memory Useful for Reliable Unanswerable Question Handling? Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27924 - Roleplaying with Structure: Synthetic Therapist-Client Conversation Generation from Questionnaires Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2510.25384 - Trajectory-Level Speculative Decoding for Diffusion Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27514 - Why Didn't It Check? Unsupported Final Claims and Their Repair in Two Tool-Equipped Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27768 - Why Knowing Both Hops Is Not Enough: Understanding Two-Hop Generalization in Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.07261 - XHotpotQA: A Benchmark for Cross-Lingual Knowledge Composition in Multi-Hop Question Answering Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27481 - Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMs Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.21751 - AI Writers Have a Consistent Stylometric Footprint, but AI Editors Do Not Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27855 - Sledgehammer or Scalpel? A Fine-grained Adaptive Framework for Implicit Hate Speech Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27462 - DiffuSent: Towards a Unified Diffusion Framework for Aspect-Based Sentiment Analysis Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.01323 - Pruning Laws for Large Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2504.04342 - LV-ROVER-MLT: Low-Resource Maltese OCR by Synthetic Fine-Tuning and Multi-Stream Arbitration Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.00250 - EvoHarmBench: Breaking Content Moderation with Iterative Human-Like Evasion Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27844 - Persuasion Index: A Theory-Guided Framework for Persuasion Analysis Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.14580 - A Probabilistic Interpretation of KV Cache Eviction Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28293 - One Form to Transfer Them All: Pretraining Multilingual Language Models Beyond Native Orthography Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.25904 - Surgical Alignment in Knowledge Graph Training for Clinical Diagnosis with Large Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26587 - Synth-JDoc: Synthesizing a Japanese Document Image Dataset for OCR with Diverse Layouts and Embedded Images Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28248 - When Stale Constraints Go Unchecked: Budgeted Verification Failures in Inherited Agent Memory Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.25553 - Representation of syntax in LLMs through the lens of linear distance and similarity-aware entropy Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27813 - A Declarative-Procedural Perspective on Expert Routing in Bilingual Mixture-of-Experts Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.15102 - RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.03889 - Diverging Transformer Predictions for Human Sentence Processing: A Comprehensive Analysis of Agreement Attraction Effects Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.16574 - MathAdv: What Theorem Provers Know, Reason, Formalize, and Generalize Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.25449 - A Survey on Rubric-Guided Reinforcement Learning for Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27505 - Benchmarking large language model agent societies against human behavioural distributions Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28182 - Phoneme- and Word-Level Metrics Using Self-Supervised Speech Representations for Forced Alignment Evaluation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28508 - SMRC: Aligning Large Language Models with Student Reasoning for Mathematical Error Correction Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.14684 - QUORUM: QUality-Optimized Routing Using Multiple annotators Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27974 - Self-Generated Text Recognition: Quality Heuristics, Cross-Task Transfer, and Downstream Bias in LLM Evaluation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26159 - Is Prosody Lost in Translation? Fine-Grained Cross-Lingual Prosody Similarity Across Languages Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27848 - SeMoCo: A Semantic-First Motion Codec for Motion Language Modeling Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.24334 - Below the Noise Floor: Bimodal Seed Collapse and Distinct Failure Modes in Small-Model Knowledge Distillation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27729 - FinExam-10K: When Retrieval Helps Financial Reasoning? Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28155 - Beyond the Rosetta Stone: Unification Forces in Generalization Dynamics Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2508.11017 - Self-Evaluation Is Already There: Eliciting Latent Judge Calibration in Base LLMs with Minimal Data Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.05122 - Tracing the complexity profiles of different linguistic phenomena through the intrinsic dimension of LLM representations Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2601.03779 - BEACON: Behavior-Anchored Cross-Source Knowledge Graph Construction for Cyber Threat Intelligence Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28394 - A Formal Limitation on Learning Human Language From Textual Corpora Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28560 - An LLM-Based Framework for Intent-Driven Network Topology Design Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.00292 - PACE: Publisher-Adaptive Content Extraction via Agentic Automation Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27466 - AIM: Anchor Identity Features, Then Match for Multimodal Large Language Model Unlearning Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28312 - TreeGraft: Adaptive Multi-Drafter Grafting for Tree-Based Speculative Decoding Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26112 - Evaluating the Performance of Large Language Models on GAOKAO Benchmark Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2305.12474 - Overview of SHROOM-Visions 2026: A Shared Task on Hallucination Detection in Large Vision-Language Models Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.25662 - From Leaky Thoughts to Private Reasoning: Controlling What LRMs Say to Themselves Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.24210 - Retrieving Relations, Detecting Fallacies: A RAG Approach to Political Debate Analysis Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27471 - OmniFusion: Simultaneous Multilingual Multimodal Translations via Modular Fusion Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2512.00234 - ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL Source: arxiv-cs-cl Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28476 - SegBench-GC: Testing Segmentation Invariance in Multi-Step Offline Goal-Conditioned Reinforcement Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27678 - Self-Explainable Multi-Label Graph Neural Network for Correlated Evidence Attribution Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27574 - Beyond Flat Netlist: Hierarchical Graph Representation Learning for Scalable Analysis of Sequential Circuits Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28188 - Ampere: Communication-Efficient and High-Accuracy Split Federated Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2507.07130 - One Model for All: Universal Pre-training for EEG based Emotion Recognition across Heterogeneous Datasets and Paradigms Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.08444 - D-TAIA: Domain-Aware LLM Adaptation for Multi-Task Predictive Process Monitoring Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28236 - Bayesian Experimental Design for Model Discrepancy Calibration: A Rivalry between Kullback--Leibler Divergence and Wasserstein Distance Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2601.16425 - Sliding-window beats linear attention Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28444 - TI$^2$PS: A Topology-Informed Inverse Design Framework for Stochastic Multicellular Pattern Formation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27931 - DARTS: Decoder-Aware Representation Tuning via Surgery for Model Merging Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28547 - Think-at-Hard: Dynamic Looped Transformers for Improved Reasoning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.08577 - CARDINAL Predicts Cardiovascular Risk From Non-contrast Cardiac CT Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27690 - Is Monte Carlo Tree Search Just Every-Visit Monte Carlo Control? Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27985 - Off the Normal Path: Learning Spatial Density Models of Node Mobility Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2411.10997 - An End-to-End Hybrid Quantum--Classical Sampling Workflow for Discrete Markov Random Fields: A Reproducible Case Study Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.09893 - Establishing Boundary KKT Convergence of Mirror Descent through Reparameterization Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.07248 - Learned Relay Representations for Forward-Thinking Discrete Diffusion Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.22967 - Deflation-PINNs: Learning Multiple Solutions for PDEs and Landau-de Gennes Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.27936 - SOMTab: Set-Order Mamba for Efficient Tabular In-Context Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27882 - Hypothesize, Evaluate, Refine: A Scientific Agent for PDE Discovery with Unknown Spatial Coefficient Fields Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27475 - SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.12015 - Post-Training VLMs for Video Mistake Detection Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28406 - Aligning Agentic World Models via Knowledgeable Experience Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2601.13247 - VICT: Verifier-Instrumented Credit Tracing for Long-Horizon LLM Agent Reinforcement Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28128 - BanglaMed-QA: A Question Answering System for Healthcare Support in Bangla Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28329 - Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNets Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2305.17010 - Comparing Classical and Quantum Machine Learning for Regression in High Energy Physics Collision Data Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28084 - Automatic Pronunciation Error Detection and Correction of the Holy Quran's Learners Using Deep Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.00094 - Biologically Inspired Mechanisms for Facilitating Grokking in Multilayer Perceptrons Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28184 - SemEnrich: Self-Supervised Semantic Enrichment of Radiology Reports for Vision-Language Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.09887 - Semantic Overlays: Mitigating Prompt Injection with Annotations Beyond Tokens and Steering Vectors Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.23873 - What Do Interaction Representations Actually Measure? Pre-Event Separability in Weakly-Supervised Violence Detection Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27879 - Var-JEPA: A Variational Formulation of the Joint-Embedding Predictive Architecture - Bridging Predictive and Generative Self-Supervised Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.20111 - SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.04208 - When Muon Meets Task Interference: A Spectral Perspective on Continual Learning and Model Merging Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27518 - Budget-Constrained Causal Bandits: Bridging Uplift Modeling and Sequential Decision-Making Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.26169 - SinkSLOT: Sinkhorn via Sparse Lifted Optimal Transport Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28262 - DART-FL: Burst-Aware Multitask Federated Learning under Dynamic Inference Demand at the Edge Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27713 - TACIT-Switch: Cost-Aware Model Escalation for LLM Agents from Censored Supervision Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27911 - Characterization of Request and Token Energy Costs for LLM Inference Workloads on GPU Platforms Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28044 - PhyMamba: Physics-Modulated Mamba for Robust Battery Health Prognostics Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27978 - Neural Regression with Embeddings for Numerical Attribute Prediction in Knowledge Graphs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26729 - Generalized Gibbs Ensemble Weighting for Forecast Combination Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28116 - RIBOSPAN: A Long-Context RNA Foundation Model for Versatile RNA Modeling Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.22849 - Robust Assortment Optimization from Observational Data Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.10696 - OpenStamp: A Watermark for Open-Source Language Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27899 - On two proofs of $d^2$ mixing of weighted Dikin walks Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28566 - Accurate prediction is not profitable advice: profit-based evaluation of machine learning nitrogen recommendations in winter wheat Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27205 - A Deeper Analysis of Block-Sparse Featurizers Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27515 - SafeStep: An Interactive Demonstration of Semantic Communication for Pedestrian Safety Monitoring Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27688 - Tensor-Accelerated Eager Multi-Resolution Grids for Evolving Large-Scale Substrates Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27612 - CURA: Certified Runtime Alarms for Computer-Use Agents Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27808 - Curvature-Aware Radius Shrinkage for Adaptive Nearest Neighbor Classification Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27634 - HARTS: Efficient Agentic Reinforcement Learning for Hybrid-Attention Models over Arbitrary Rollout Trees Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28158 - Explainable Uncertainty Estimation for Reliable Medical AI Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28052 - How Do Linear Probes Emerge? A Circuit-Tracing Framework with Concept-Targeted Attribution Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27510 - Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.26895 - Multiscale Community-Based Fingerprinting of Signed Functional Networks Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27483 - Explainable Diabetic Retinopathy Classification Using Vision Foundation Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28207 - Towards a mathematical theory of superposition Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27540 - Residual-Guided Randomized Neural Networks Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28267 - Learning to Difference: Adaptive Reversible Differencing (AdaRDiff) for Time Series Forecasting Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28134 - There and Back Again: Bidirectional Diffusion Bridges for Multimodality Translation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27885 - Trajectory balance: Improved credit assignment in GFlowNets Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2201.13259 - Accelerating LLM Inference via Vector Index Based Output Embeddings Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27460 - Beyond Procrustes distances: a multilinear Gromov-Wasserstein distance capturing chirality Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27774 - How Architecture and Training Affect TPC Representations Across Experiments Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.21756 - Agentic-Kube: A Graph-Enhanced Multi-Agent Reinforcement Learning Framework for Multi-Objective Kubernetes Scheduling Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.12031 - An algebraic proof of Colombo's difference-power determinant conjecture Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28274 - Blog: Survey of Optimizers Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28557 - Beyond Pairwise Graphs in Science: Hypergraph Adaptive Wavelet Operators for Parametric PDEs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27883 - SymboLLM-FE: LLM-Accelerated Symbolic Regression for Automated Feature Engineering on Tabular Data Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28408 - Actionable CBFI: Integrating Structural Decomposition and Causal Counterfactual Recourse for Tabular Machine Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27821 - Anchored Scenario Coverage for Failure-Aware First-Hit Batch Inverse Design Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27873 - Real-Time Monitoring of MHD Liquid Metal Flows with Shallow Recurrent Decoders Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28366 - WINO: A Weak-Form Physics Informed Neural Operator for Hyperelasticity on Variable Domains Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.24651 - Node-wise Feature Encoding for Neural Performance Prediction Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27794 - Generalized Splines and Gaussian Processes Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28446 - GAN-Diff : Coupling Pretrained WGAN-GP Features with Conditional Diffusion U-Nets Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.22272 - ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Space Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.27443 - Online Learning-to-Defer with Varying Experts Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.12340 - D3-Gym: Constructing Real-World Verifiable Environments for Data-Driven Discovery Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.27977 - PolicyLong: Towards On-Policy Context Extension Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.07809 - CheXtriev: Anatomy-Centered Representation for Case-Based Retrieval of Chest Radiographs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28137 - Exact Risk Ratios for Weighted Data Selection in Linear Regression Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28007 - Landau theory of quenched criticality in linear in-context learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28059 - Joint Bayesian Inference of Graphical Structure and Parameters with a Single Generative Flow Network Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2305.19366 - Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spots Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27782 - ToolSense: A Diagnostic Framework for Auditing Parametric Tool Knowledge in LLMs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.12451 - LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysis Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.30434 - Marginal Coverage Credit Reduces Redundant Exploration in Parallel State-Entropy Optimization Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27507 - When Can Conditional Flow Matching Replace Pointwise Negative Log-Likelihood? Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28010 - Mine and Refine: Optimizing Graded Relevance in E-commerce Semantic Search Retrieval Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.17654 - TokenPilot: Cache-Efficient Context Management for LLM Agents Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.17016 - Depth-Aware Pothole Detection Using YOLO and RT-DETR at the Edge Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27633 - SpecGradFilter: A Spectral Gradient Filtering Framework for Taming Federated Heterogeneity Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.04189 - REPLICANT: Learning Policies for Evading and Hardening Malware Detectors Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28499 - What Neural Network Field Theory Can and Cannot Realise on a Computer Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.21523 - SciReC: Diagnostic Evaluation of Multimodal, Multi-Turn Relational Reasoning with Adaptive Interaction Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27461 - Prompts Without Evidence: How Neuroimaging Mentions Shift Clinical Vision-Language Model Predictions Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.28387 - Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPO Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27351 - Towards Large-Scale Heterogeneous Data Organization for Scientific Foundation Models: A Nuclear Fusion Case Study Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27578 - Unsupervised Continual Learning with Growing Self-Organizing Maps and Synthetic Replay Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27662 - Acquire, Repair, Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agents Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28458 - The Autonomy Tax: Defense Training Breaks LLM Agents Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.19423 - Performative Privacy: When Differential Privacy Maximizes Utility Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28198 - Not to Break, but to Attest: Adversarial Probes for Privacy-Preserving LLM Verification Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27954 - RecourseBench: A Modular Framework for Reproducible Algorithmic Recourse Evaluation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.16113 - Amortizing intractable inference in diffusion models for vision, language, and control Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2405.20971 - Deriving Scaling Laws for OpenEuroLLM Models: Learning Rate, Batch Size and Loss Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28308 - Meta-Prompt Optimization for LLM-Based Sequential Decision Making Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2502.00728 - Robust Chance-Constrained Optimization using a Continuous Parameter Space Wasserstein-2 Ambiguity Set of Gaussian Mixtures Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.17018 - G-Loss: Graph-Guided Fine-Tuning of Language Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.25853 - On-policy Distillation with Verifiable Reward Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.24696 - Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gap Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27512 - Multilingual Lexical Feature Analysis of Spoken Language for Predicting Major Depression Symptom Severity Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.07011 - On the Computational and Statistical Efficiency of the Empirical Maximum Entropy on the Mean Method Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27705 - Improved off-policy training of diffusion samplers Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2402.05098 - Euclidean Fourier Neural Operators Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28425 - Dandelion: A Spherical Flower for Neural Simulation of Planetary Dynamics Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27521 - Physics-informed learning for the inverse problem in resonant ultrasound spectroscopy Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27590 - Efficient Online Continual Foundation Model Fine-Tuning for Predictive Process Monitoring Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28237 - Beyond Non-IID: Learner--Client Distribution Mismatch in Federated Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27715 - Curvature-Conditioned Multiscale Momentum with Sphere Constraints for LLM Pretraining Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28442 - How Far Should Tokenization Go? Predictive Effectiveness and Relational Losslessness Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.18025 - Rethinking Speaker Embeddings for Speech Generation: Sub-Center Modeling for Capturing Intra-Speaker Diversity Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2407.04291 - TraceML: An Empirical Analysis of Human-Agent Planning in Machine Learning Development Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.26086 - QGPINNs: A Physics-Informed Neural Network Framework for Nonlocal Differential Equations on Quantum Graphs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28589 - Prequential posteriors Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.17721 - The Calls are Coming from Inside the Model: Investigating Probe-based Detection of Tool-Calling Errors in LLMs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27750 - ED-CSP: Crystal Structure Prediction from Electron Diffraction Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.06448 - FlowCorrect: Efficient Interactive Correction of Generative Flow Policies for Robotic Manipulation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.22056 - Conditional Diffusion Models for Energy-Efficient Driving Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28142 - Destroy Me: Automatic Artifact Generation for Histopathology Images Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27516 - DiffAnon: Diffusion-based Prosody Control for Voice Anonymization Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.26281 - InfoMamba: An Attention-Free Hybrid Mamba-Transformer Model Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.18031 - GREAT: Generalizable Backdoor Attacks in RLHF via Emotion-Aware Trigger Synthesis Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2510.09260 - Ab initio Modeling of MoS2/Oxide Device Interfaces with Machine Learned Electronic Structures Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27533 - Advancing Interaction-Sensitive Feature Selection: Novel Relief-Based Algorithms, Expanded Comparisons, and Recommendations for Biomedical Data Mining Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28552 - Fast Weight Attention for Continual Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27763 - Mixture of Multicenter Experts in Multimodal AI for Debiased Radiotherapy Target Delineation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2410.00046 - RiskBlend: A Multi-Signal Framework for Test Input Prioritization in Machine Learning Regression Testing Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27704 - Quantum Federated Learning Based on Bures--Uhlmann Geometry for Heterogeneous Noisy Clients Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28379 - Attention as Conditioning: What Classical Learning Theory Predicts About Linear Transformers Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2508.08289 - Initialization Is Critical: Advancing Federated Short-Term Load Forecasting under Load Heterogeneity via Model Initialization Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27791 - Trust the Mass: Forced Weights in KV-Cache Eviction Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.25230 - I-FLOP: Fast Learning of Order and Parents from Interventional Data Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28245 - Empowering Local Agriculture: A Deep Learning-Powered Web System for Identifying Bangladeshi Mango Varieties Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28161 - Leveraging a Foundation Model for the EEG-Based Diagnosis of Alzheimer's Disease Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27719 - Can Subgraph Explanations Be Weaponized to Steal Graph Neural Networks? Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.30470 - Camera-Agnostic Pruning of 3D Gaussian Splats via Descriptor-Based Beta Evidence Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.21933 - ERP-XTTN: Interpretable Prototype-Guided Cross-Attention for Cross-Subject ERP Classification Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.02939 - Aspiration-based Perturbed Learning Automata in Games with Noisy Utility Measurements. Part A: Stochastic Stability in Non-zero-Sum Games Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2511.11602 - Quantum SEDONet: Spectrally-Embedded Quantum Deep Operator Networks for Partial Differential Equations Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27626 - Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representations Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28092 - Spectral Features Dominate BCG Respiratory-Event Detection: A Large-Scale Patient-Independent Comparison of Feature Groups in Sleep Apnea Patients Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28242 - JEPA-x: Cross-Predictive Physics Grounding for Forecastable Latent Dynamics Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.24044 - Probabilistic Symbolic Regression for Equation Discovery via Operator-induced and Regularized Symbolic Forests Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.19710 - Transformer-Based Autonomous Driving Models and Deployment-Oriented Compression: A Survey Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2304.10891 - Simplex-to-Euclidean Bijection for Conjugate and Calibrated Multiclass Gaussian Process Classification Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2603.16621 - Personalized and Multi-View Representation for Federated Cold-Start Recommendation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27826 - Emergent aggregation from collective foraging Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28046 - More Data Cannot Break a Symmetry: Identifiability by Design Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27651 - The Approximation Rank of Softmax Attention: Sharp Geometric Laws and Robust Interaction Dimension Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28150 - Timing-Aware Repurchase Prediction for Web-Scale E-Commerce: Survival Models for Multi-Surface Grocery Recommendation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28393 - Shift Before You Learn: Enabling Low-Rank Representations in Reinforcement Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.05193 - Conformal Risk-Averse Decision Making with Optimized Certainty Equivalent Risk Control Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28179 - Where Steering Signals Come From: Activation Source Selection in Activation Steering Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.25270 - Effectiveness of IoT and Deep Learning for Detection and Severity Assessment of Postelectrotermes militaris in Tea Plantations Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27480 - Beyond Output Correctness: Benchmarking and Evaluating Large Language Model Reasoning in Coding Tasks Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2604.12379 - Large Reasoning Models Learn Better Alignment from Flawed Thinking Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2510.00938 - OceanGym: A Benchmark Environment for Underwater Embodied Agents Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.26536 - Diffusion models as plug-and-play priors Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2206.09012 - How Proper Scoring Rules Shape LLM Forecasting Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28482 - Under-Mattress Temporal Sensing for Next-Day Agitation Risk Scoring in Dementia Wards Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28152 - Closing the Operational Gap in Semantic Caching Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.19719 - Learning to Transfer Across Modes: Towards Unified Urban Mobility Forecasting Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28273 - The Instability of Safety: How Random Seeds and Temperature Expose Inconsistent LLM Refusal Behavior Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2512.12066 - Twin Worlds: Equivariance-Based Abstention for Evidence-Grounded Reasoning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28018 - JuryProbe: An Empirical Consensus-Risk Diagnostic for Routing Reference-Free Factuality Judge Panels to Grounded Verification Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.20607 - Diffusion Distillation for Efficient Weather Ensembles Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27728 - Beyond Search-Imitation: Prior-Directed Exploration for Searchless Chess Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27757 - DAMP: Decay-Aware Mixed-Precision Recurrent-State Quantization Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27513 - Generalized Context in Cross Attention for Transfer Learning of Disjoint Tabular Data Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28209 - The Discrete-Log Clock: How a Transformer Learns Modular Multiplication Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2606.17399 - Temporal Memory-Aware Online Test-Time Adaptation on Dynamic Graphs Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27948 - Locked Evaluation Surfaces: Transfer Failure and Sampling-Depth Entanglement in CRISPRi Perturbation-Effect Prediction Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.00152 - RegCL: Compact Continual SAM Adaptation for Visual Grounding in Multi-Sensorial Media Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2507.12297 - Examining the robustness of Physics-Informed Neural Networks to noise for Inverse Problems Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2509.20191 - Learning Fast Monomial Orders for Gr\"obner Basis Computations Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2602.02972 - Learning between the peaks: sharp asymptotics for kernel ridge regression under power-law anisotropy Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28564 - Recirculation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.17981 - ImplicitTerrainV2: Wavelet-Guided Spatially Adaptive Neural Terrain Representation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.22556 - Learning a Size-Weight Frontier for Synthetic-Augmented Inference Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28576 - Stay Within Your Bounds: Distance-Guided Decoding for Guaranteed Context-Free Grammar Compliance Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28229 - Class Incremental Continual Learning with Self-Organizing Maps and Synthetic Replay Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2508.21240 - GRACE:Gradient-guided Coreset Selection for LLM Unlearning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28361 - A Method for Layer Bit-Width Allocation in LLM Quantization via Performance Maximization Under a Quality-Degradation Constraint Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28003 - FedEHR-Agents: Federated Agentic Optimization for Automated EHR Modeling Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27856 - On the Depth Scalability of Logic Gate Networks Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2607.21633 - RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requests Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27831 - Optimal Transport for Network Comparison: A Review with Machine Learning Applications Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.27500 - Parser States Already Know: Structure-Conditioned KV Persistence for Structured Generation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28276 - An Enclosed Mode Is a Gauge Choice: Topology Relative to Reach in Certified Code World Models Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28541 - Aero Hand Open: A Simulation-Ready Tendon-Driven Hand for Dexterous Manipulation Learning Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28578 - VISTA: Verifier-Informed Student-to-Teacher Adaptation for On-Policy Self-Distillation Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28306 - More Expressive Feedforward Layers: Part I. Token-Adaptive Mixing of Activations Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2605.26647 - Biases in Expected Goals Models Confound Finishing Ability Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2401.09940 - Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detection Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28375 - EXPOSE: Explainable and Domain-Robust Embeddings from Pathology Vision Foundation Models using Sparse Autoencoders Source: arxiv-cs-lg Topic: AI Research (+AI) URL: https://arxiv.org/abs/2608.28191 ## Community - Introducing wrapture Source: simon-willison Topic: Python (+AI, Django) URL: https://simonwillison.net/2026/Aug/31/introducing-wrapture/ - Quoting Andrew Digby Source: simon-willison Topic: Python (+AI, Django) URL: https://simonwillison.net/2026/Aug/31/andrew-digby/ - Ed Crewe: From Routing Checks to Trajectory Testing: Evaluating an Agentic Chatbot Source: planet-python Topic: Python (+AI, PostgreSQL) URL: http://edcrewe.blogspot.com/2026/08/from-routing-checks-to-trajectory.html - Security scanning my own code with Scrutineer and local coding models Source: ocaml-planet Topic: OCaml (+AI) URL: https://anil.recoil.org/notes/scrutineer-local-llm