Every paper checked — nothing hidden

All papers

Every paper checked for brief #4 of 12–19 Oct 2026 — nothing is hidden. Brief papers first.

All17,313 In brief24 More (on-topic)3,333 Not relevant13,754 Not assessed87 Older352 Could not confirm93 Blogs with no paper266 All blogs284

Paper	Topic	Relevance
MedEvo: Medical Reasoning Evolution through Knowledge Navigation Proposes MedEvo, an iterative training framework that repairs multimodal medical reasoning trajectories using minimal hints and taxonomy-guided feedback. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on medical multimodal fine-tuning, not autonomous agents.
MedEvo: Medical Reasoning Evolution through Knowledge Navigation Proposes MedEvo, an iterative training framework that repairs multimodal medical reasoning trajectories using minimal hints and taxonomy-guided feedback. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on diagnostic task accuracy rather than safety guardrails or reliability.
HelpMeOut: A Neuropsychology-Grounded Benchmark for Error-Aware Monitoring of Household Task Execution Introduces HelpMeOut, an embodied AI benchmark for diagnosing human execution errors across household activity trajectories motivated by neuropsychology research. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Focuses on household assistance monitoring, not industrial defect inspection.
zPocket: Better Pocket Finding with Frozen Co-Folding Features Introduces the PocketWatch benchmark and zPocket, an approach training pocket-prediction heads on frozen representations from protein co-folding models. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on protein-ligand binding pockets, not LLM agents.
zPocket: Better Pocket Finding with Frozen Co-Folding Features Introduces the PocketWatch benchmark and zPocket, an approach training pocket-prediction heads on frozen representations from protein co-folding models. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on computational drug discovery rather than AI deployment safety.
zPocket: Better Pocket Finding with Frozen Co-Folding Features Introduces the PocketWatch benchmark and zPocket, an approach training pocket-prediction heads on frozen representations from protein co-folding models. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Evaluates biological pocket prediction, not LLM benchmarks.
zPocket: Better Pocket Finding with Frozen Co-Folding Features Introduces the PocketWatch benchmark and zPocket, an approach training pocket-prediction heads on frozen representations from protein co-folding models. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on structural biology models, not model compression.
LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification Presents LeanPRM, a method using the Lean proof assistant to generate step-level correctness labels for training natural-language mathematical proof verifiers. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on mathematical proof step-verification, not multi-step agent actions.
LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification Presents LeanPRM, a method using the Lean proof assistant to generate step-level correctness labels for training natural-language mathematical proof verifiers. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on automated supervision for verifiers, not model compression.
LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification Presents LeanPRM, a method using the Lean proof assistant to generate step-level correctness labels for training natural-language mathematical proof verifiers. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Addresses proof verification accuracy rather than deployment safety.
Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity Studies the trade-off between spatial smoothness and gene-level fidelity when mapping single-cell RNA sequencing profiles to images for neural networks. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on biological representation learning, not model compression.
Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity Studies the trade-off between spatial smoothness and gene-level fidelity when mapping single-cell RNA sequencing profiles to images for neural networks. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on genomics and RNA representations, not LLM agents.
Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity Studies the trade-off between spatial smoothness and gene-level fidelity when mapping single-cell RNA sequencing profiles to images for neural networks. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Studies gene-to-image encoding, not LLM benchmark evaluation.
Understanding Gene Expression-to-Image Learning: Spatial Structure and Information Fidelity Studies the trade-off between spatial smoothness and gene-level fidelity when mapping single-cell RNA sequencing profiles to images for neural networks. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Concerns biological data representations rather than AI safety.
A Node Embedding Framework for Model-Centric Conformance Checking in Process Mining Introduces a graph embedding and Procrustes alignment framework to provide a bounded metric for evaluating discovered process models against ground truth. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Presents process mining graph methods, not LLM agents.
A Node Embedding Framework for Model-Centric Conformance Checking in Process Mining Introduces a graph embedding and Procrustes alignment framework to provide a bounded metric for evaluating discovered process models against ground truth. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Addresses process mining metrics rather than AI safety or reliability.
A Node Embedding Framework for Model-Centric Conformance Checking in Process Mining Introduces a graph embedding and Procrustes alignment framework to provide a bounded metric for evaluating discovered process models against ground truth. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Evaluates business process models, not language models.
From Token Correction to Claim Admission: Training-Free Hallucination Mitigation for Large Vision-Language Models Proposes a training-free framework that mitigates vision-language model hallucinations by admitting claims only when supported relative to a visual-null input. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Studies general vision-language hallucination, not industrial defect detection.
From Token Correction to Claim Admission: Training-Free Hallucination Mitigation for Large Vision-Language Models Proposes a training-free framework that mitigates vision-language model hallucinations by admitting claims only when supported relative to a visual-null input. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on vision-language hallucination, not tool-using agents.
From Token Correction to Claim Admission: Training-Free Hallucination Mitigation for Large Vision-Language Models Proposes a training-free framework that mitigates vision-language model hallucinations by admitting claims only when supported relative to a visual-null input. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Proposes a hallucination mitigation algorithm rather than a new benchmark.
Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning Develops SCRIM, a method quantifying state criticality in imitation learning by measuring downstream value degradation under bounded action perturbations. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents an imitation learning method, not an LLM benchmark.
Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning Develops SCRIM, a method quantifying state criticality in imitation learning by measuring downstream value degradation under bounded action perturbations. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Addresses general imitation learning theory rather than LLM-based agents.
Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning Develops SCRIM, a method quantifying state criticality in imitation learning by measuring downstream value degradation under bounded action perturbations. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on policy optimization in imitation learning rather than AI safety.
Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning Develops SCRIM, a method quantifying state criticality in imitation learning by measuring downstream value degradation under bounded action perturbations. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Applies to general robotics and control, not industrial vision inspection.
Where Small Mistakes Matter: Quantifying State Criticality for Imitation Learning Develops SCRIM, a method quantifying state criticality in imitation learning by measuring downstream value degradation under bounded action perturbations. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on demonstration curation in imitation learning, not model efficiency.
Adaptive-Group Process Rewards for GUI Agents Introduces an adaptive-group process reward framework that partitions GUI trajectories into subgoals to provide intermediate credit assignment for reinforcement learning. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Develops a training reward framework rather than an evaluation benchmark.
Adaptive-Group Process Rewards for GUI Agents Introduces an adaptive-group process reward framework that partitions GUI trajectories into subgoals to provide intermediate credit assignment for reinforcement learning. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Targets policy optimization efficiency rather than safety or reliability.
ECG-RSI: Recursive Self-Improvement through Targeted Electrocardiogram Waveform Evolution Proposes a recursive self-improvement framework for electrocardiogram analysis models combining verified synthetic waveform edits with targeted intervention selection. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Studies recursive self-improvement on ECGs, not general LLM evaluation.
ECG-RSI: Recursive Self-Improvement through Targeted Electrocardiogram Waveform Evolution Proposes a recursive self-improvement framework for electrocardiogram analysis models combining verified synthetic waveform edits with targeted intervention selection. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on domain self-improvement, not model compression or efficient serving.
ECG-RSI: Recursive Self-Improvement through Targeted Electrocardiogram Waveform Evolution Proposes a recursive self-improvement framework for electrocardiogram analysis models combining verified synthetic waveform edits with targeted intervention selection. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on specialized biomedical signal learning rather than deployment safety.
ECG-RSI: Recursive Self-Improvement through Targeted Electrocardiogram Waveform Evolution Proposes a recursive self-improvement framework for electrocardiogram analysis models combining verified synthetic waveform edits with targeted intervention selection. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on ECG waveform classification rather than agent systems.
Keep Your Receipts: Trustworthy Cleanup of Agent-Built Knowledge Graphs Presents PolyGraph, a method storing provenance receipts during knowledge graph construction to enable precise updates when source documents are retracted. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on knowledge graph maintenance, not LLM benchmark evaluation.
Upstream Influence, Downstream Re-Anchoring in LLM Agent Chains Analyzes how multi-agent workflow architecture affects the propagation and attenuation of coordinated stance manipulation across sequential agent chains. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Studies belief propagation rather than LLM benchmarking methodologies.
Evidence-First Agentic Verification Introduces Sauron, an evidence-first agentic verifier that checks both environment state and execution traces against original sources to validate task rubrics. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on verifier accuracy and reward hacking, not efficient serving.
TSRL: Trigger-based Supervised Reinforcement Learning for Continual Agent Training This paper proposes a training framework that triggers supervised fine-tuning specifically on failure states during reinforcement learning to prevent agent forgetting. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses continual learning degradation, not model compression or efficiency.
TSRL: Trigger-based Supervised Reinforcement Learning for Continual Agent Training This paper proposes a training framework that triggers supervised fine-tuning specifically on failure states during reinforcement learning to prevent agent forgetting. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Evaluates on benchmarks but proposes a training framework, not evaluation.
TSRL: Trigger-based Supervised Reinforcement Learning for Continual Agent Training This paper proposes a training framework that triggers supervised fine-tuning specifically on failure states during reinforcement learning to prevent agent forgetting. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on capability retention rather than safety guardrails or alignment.
SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models This paper presents a compact decision model designed to screen malicious agent tool skills with high recall directly from specification files. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on malicious skill detection rather than general benchmark evaluation.
SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models This paper presents a compact decision model designed to screen malicious agent tool skills with high recall directly from specification files. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Does not address retrieval-augmented generation or document search.
SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models This paper presents a compact decision model designed to screen malicious agent tool skills with high recall directly from specification files. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Focuses on screening malicious skills rather than software engineering tasks.
SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models This paper presents a compact decision model designed to screen malicious agent tool skills with high recall directly from specification files. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Does not involve multimodal inputs like image, video, or audio.
SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models This paper presents a compact decision model designed to screen malicious agent tool skills with high recall directly from specification files. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Presents a compact classifier, but efficiency is not the primary research focus.
Safety for Whom? Benchmarking Understanding Beyond Refusal in Egocentric Multimodal Assistants This paper introduces a benchmark to evaluate whether egocentric multimodal assistants understand who is harmed rather than solely tracking output refusal. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Studies multimodal assistant safety rather than agent planning or tools.
CoCA: Criterion-Aware Credit Assignment for Rubric-Based Reinforcement Learning This paper proposes a credit-assignment method that uses rubric evaluation outcomes to allocate token-level advantages during reinforcement learning. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Addresses RL credit assignment, not multi-step agent planning.
CoCA: Criterion-Aware Credit Assignment for Rubric-Based Reinforcement Learning This paper proposes a credit-assignment method that uses rubric evaluation outcomes to allocate token-level advantages during reinforcement learning. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on post-training RL optimization rather than benchmark evaluation.
CoCA: Criterion-Aware Credit Assignment for Rubric-Based Reinforcement Learning This paper proposes a credit-assignment method that uses rubric evaluation outcomes to allocate token-level advantages during reinforcement learning. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on reward optimization rather than model compression or inference efficiency.
CoCA: Criterion-Aware Credit Assignment for Rubric-Based Reinforcement Learning This paper proposes a credit-assignment method that uses rubric evaluation outcomes to allocate token-level advantages during reinforcement learning. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Concerns general RL training rather than safety alignment or guardrails.
Explore with Memory, Learn without It: In-Context Exploration and Self-Distillation for LLM Reasoning This paper presents a reinforcement learning approach that uses in-context dynamic memory during rollouts and self-distillation to accelerate reasoning exploration. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Targets reasoning efficiency rather than safety or reliability risks.
Explore with Memory, Learn without It: In-Context Exploration and Self-Distillation for LLM Reasoning This paper presents a reinforcement learning approach that uses in-context dynamic memory during rollouts and self-distillation to accelerate reasoning exploration. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on RLVR exploration and self-distillation rather than agent workflows.
Explore with Memory, Learn without It: In-Context Exploration and Self-Distillation for LLM Reasoning This paper presents a reinforcement learning approach that uses in-context dynamic memory during rollouts and self-distillation to accelerate reasoning exploration. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents a training framework rather than a testing benchmark.
Train for Many, Update with One: Improving Multi-Sample Reasoning via Fenchel-Bregman Posterior Learning This paper introduces a posterior learning approach that improves multi-sample reasoning coverage by updating models using single posterior draws. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on optimization efficiency, not safety or guardrails.
Train for Many, Update with One: Improving Multi-Sample Reasoning via Fenchel-Bregman Posterior Learning This paper introduces a posterior learning approach that improves multi-sample reasoning coverage by updating models using single posterior draws. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Proposes an optimization method rather than evaluation methodologies or benchmarks.
Train for Many, Update with One: Improving Multi-Sample Reasoning via Fenchel-Bregman Posterior Learning This paper introduces a posterior learning approach that improves multi-sample reasoning coverage by updating models using single posterior draws. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on multi-sample posterior optimization, not multi-step agent systems.
DuplexAct: How Interactive Are Frontier Full-Duplex Spoken Language Models? This paper presents a benchmark with an action taxonomy to evaluate the interactive conversation capabilities of full-duplex spoken language models. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on spoken conversational turn-taking, not tool-using or planning agents.
DuplexAct: How Interactive Are Frontier Full-Duplex Spoken Language Models? This paper presents a benchmark with an action taxonomy to evaluate the interactive conversation capabilities of full-duplex spoken language models. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Evaluates conversational flow rather than AI safety or reliability.
HiSTOR: Structured Treatment-outcome Representation Learning for High-Dimensional Longitudinal Data This paper proposes an end-to-end framework integrating nonlinear dimension reduction with longitudinal counterfactual prediction for high-dimensional observational data. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Covers causal treatment estimation rather than AI deployment safety or guardrails.
HiSTOR: Structured Treatment-outcome Representation Learning for High-Dimensional Longitudinal Data This paper proposes an end-to-end framework integrating nonlinear dimension reduction with longitudinal counterfactual prediction for high-dimensional observational data. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses statistical covariate dimension reduction, not LLM compression or serving.
HiSTOR: Structured Treatment-outcome Representation Learning for High-Dimensional Longitudinal Data This paper proposes an end-to-end framework integrating nonlinear dimension reduction with longitudinal counterfactual prediction for high-dimensional observational data. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on counterfactual prediction methods rather than LLM evaluation benchmarks.
HiSTOR: Structured Treatment-outcome Representation Learning for High-Dimensional Longitudinal Data This paper proposes an end-to-end framework integrating nonlinear dimension reduction with longitudinal counterfactual prediction for high-dimensional observational data. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Deals with causal inference for longitudinal data, not LLM agents.
Coupled Harness--Model Evolution for Scientific Equation Discovery This paper presents a coupled evolution framework that co-adapts an executable tool library and a model's policy for scientific equation discovery. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Introduces a tool-learning method rather than a general LLM evaluation benchmark.

13,561–13,620 of 17,313← Previous Next →