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
PRISM: Transfer and Limits of Activation Interventions in Tool-Calling Language Models Studies the transferability of internal activation interventions to reduce unsafe tool execution across model precisions and languages. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Proposes an alignment framework rather than general LLM evaluation benchmarks.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Focuses on RL training algorithms rather than deployment safety or guardrails.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Does not specifically study code generation or software engineering tasks.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not investigate retrieval-augmented generation or database retrieval.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Focuses on RL optimization rather than small models or efficient inference.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Does not study multimodal models combining text with vision or audio.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Proposes a training algorithm rather than an LLM evaluation benchmark.
Gibbs-Gap Policy Optimization for Off-Policy RL with Verifiable Rewards Introduces an off-policy reinforcement learning algorithm for language models using verifiable rewards without importance sampling weights. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Focuses on RL optimization theory rather than agentic planning or tools.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not study AI safety, jailbreaking, or deployment guardrails.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not investigate retrieval-augmented generation pipelines.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Does not evaluate autonomous tool-using agents or multi-agent planning.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about code generation or software engineering tasks.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Focuses on speech audio processing rather than multimodal cross-modality models.
PEDBench: Can Speech Models Detect Phonetic Errors? Introduces a benchmark combining natural, scripted, and synthetic audio to evaluate phonetic error detection in speech models. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Does not evaluate model compression, distillation, or serving efficiency.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about code generation, coding assistants, or program repair.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Not about autonomous agents or multi-step agent planning.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not examine retrieval-augmented generation systems.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not address AI safety, robustness, or reliability guardrails.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Studies morphological tokenization rather than model compression or inference efficiency.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Focuses on linguistic tokenization, not LLM benchmark evaluation.
Does Morphological Correctness Matter for Tokenization? A Scrambled Control Study in Tagalog Tests whether linguistically accurate morpheme boundaries in tokenizers improve translation performance compared to scrambled control segmentations. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Focuses strictly on text tokenization without multimodal data.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about code generation, repair, or programming assistants.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not investigate retrieval-augmented generation or document search.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Does not focus on general LLM evaluation or benchmarks.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not address AI safety, jailbreaks, or deployment guardrails.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Does not investigate model compression, distillation, or efficient serving.
CogDrift-4D: Hierarchical Multimodal Modeling of Temporal Behavioral Decay for MCI Detection Proposes a hierarchical multimodal transformer to model long-term behavioral changes across multi-session interactions for cognitive impairment detection. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Not about autonomous multi-step planning or tool-using agents.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Not focused on general LLM evaluation benchmarks or methodologies.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not study external document retrieval or grounding architectures.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not focus on AI safety, alignment, or guardrails.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Does not address model efficiency, quantization, or serving compression.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Does not address code generation, repair, or software engineering.
Task-Centric Multimodal Learning for Cognitive Impairment Detection Evaluates a multimodal screening method using voice, video, and text from voice-assistant interactions to detect cognitive impairment. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Not focused on autonomous multi-step planning or tool use agents.
Prune the Summary, Not the History: Source-Judged Compaction for Long-Horizon Agents Introduces a method to filter unsupported compressor inferences from long-horizon agent summaries to avoid compounding execution errors. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Focuses on agent history compaction rather than external knowledge retrieval.
Prune the Summary, Not the History: Source-Judged Compaction for Long-Horizon Agents Introduces a method to filter unsupported compressor inferences from long-horizon agent summaries to avoid compounding execution errors. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Operates on textual agent interaction history without multimodal components.
Prune the Summary, Not the History: Source-Judged Compaction for Long-Horizon Agents Introduces a method to filter unsupported compressor inferences from long-horizon agent summaries to avoid compounding execution errors. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Proposes a compaction technique rather than an LLM evaluation benchmark.
Prune the Summary, Not the History: Source-Judged Compaction for Long-Horizon Agents Introduces a method to filter unsupported compressor inferences from long-horizon agent summaries to avoid compounding execution errors. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about software development, programming tools, or code repair.
VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents Presents a benchmark with human-annotated visual evidence to evaluate retrieval and question answering across long documents. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Does not cover model compression, distillation, or efficient inference serving.
VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents Presents a benchmark with human-annotated visual evidence to evaluate retrieval and question answering across long documents. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not address deployment safety, jailbreaks, or red-teaming.
VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents Presents a benchmark with human-annotated visual evidence to evaluate retrieval and question answering across long documents. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Presents a QA dataset rather than multi-step autonomous agent architectures.
VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents Presents a benchmark with human-annotated visual evidence to evaluate retrieval and question answering across long documents. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not related to code generation, software repair, or programming tasks.
VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents Presents a benchmark with human-annotated visual evidence to evaluate retrieval and question answering across long documents. Preprint	Computer vision for industrial applications via arxiv · 17 Oct, 06:17	Not relevant Analyzes documents and PDFs rather than industrial defect inspection.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Focuses on multiple-choice evaluation metrics, not autonomous planning agents.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about code generation, software development, or code evaluation.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Examines text multiple-choice questions without multimodal data.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Not about model compression, distillation, or serving efficiency.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Focuses on benchmark evaluation artifacts rather than deployment safety guardrails.
One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation Deconstructs multiple-choice option bias in language models into two factors tracking answer flips and positional skew. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not concern retrieval-augmented generation or grounding.
Can LLMs Anticipate Their Cruxes? Explores methods to extract the underlying decisive sub-questions that drive language model decisions across diverse generated outputs. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Analyzes decision rationale distributions rather than autonomous agent systems.
Can LLMs Anticipate Their Cruxes? Explores methods to extract the underlying decisive sub-questions that drive language model decisions across diverse generated outputs. Preprint	Small and efficient models via arxiv · 17 Oct, 06:17	Not relevant Does not explore model quantization, pruning, or small models.
Can LLMs Anticipate Their Cruxes? Explores methods to extract the underlying decisive sub-questions that drive language model decisions across diverse generated outputs. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not related to code generation, program repair, or software engineering.
Can LLMs Anticipate Their Cruxes? Explores methods to extract the underlying decisive sub-questions that drive language model decisions across diverse generated outputs. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Studies text reasoning distributions without multimodal elements.
Can LLMs Anticipate Their Cruxes? Explores methods to extract the underlying decisive sub-questions that drive language model decisions across diverse generated outputs. Preprint	Retrieval-augmented generation (RAG) via arxiv · 17 Oct, 06:17	Not relevant Does not involve external document retrieval or RAG architectures.
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Applies strictly to text retrieval without multimodal data.
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Preprint	AI safety and reliability in deployment via arxiv · 17 Oct, 06:17	Not relevant Does not address safety alignment, red teaming, or guardrails.
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Preprint	LLM evaluation and benchmarks via arxiv · 17 Oct, 06:17	Not relevant Proposes a retrieval method rather than a general evaluation benchmark.
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Not about software development, coding tools, or code generation.
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval Proposes an adaptive reasoning framework that supervises intermediate retrieval states and stops reasoning steps dynamically to cut latency. Preprint	AI agents via arxiv · 17 Oct, 06:17	Not relevant Focuses on retriever reasoning rather than autonomous agent tool use.
INCONVENIENT: A Benchmark to Assess LLM Generalization Capabilities from Convenience Samples Introduces a benchmark to test model generalization under systematic convenience sampling biases in tabular data. Preprint	LLMs for code via arxiv · 17 Oct, 06:17	Not relevant Focuses on tabular prediction benchmarks rather than code generation.
INCONVENIENT: A Benchmark to Assess LLM Generalization Capabilities from Convenience Samples Introduces a benchmark to test model generalization under systematic convenience sampling biases in tabular data. Preprint	Multimodal models via arxiv · 17 Oct, 06:17	Not relevant Covers tabular data prediction rather than multimodal vision-language models.

4,381–4,440 of 17,313← Previous Next →