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
When Is Smaller Greener? Right-Sizing Language Models for Agentic AI The paper presents a decision model to evaluate language model efficiency based on energy and cost per completed agentic task. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Proposes a sizing decision model rather than an evaluation benchmark.
Asymptotic Behavior of Transformers The paper mathematically analyzes long-sequence generation in transformers, establishing temperature thresholds required for internal cache convergence. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on asymptotic convergence properties, not model compression or serving.
Asymptotic Behavior of Transformers The paper mathematically analyzes long-sequence generation in transformers, establishing temperature thresholds required for internal cache convergence. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Mathematical proof of convergence properties, not an empirical benchmark.
Asymptotic Behavior of Transformers The paper mathematically analyzes long-sequence generation in transformers, establishing temperature thresholds required for internal cache convergence. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Mathematical analysis of attention dynamics, not autonomous agent systems.
Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering The paper introduces a test showing that steering vectors designed for competing ethical frameworks behave interchangeably in open-weight models. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Studies activation steering vectors, not autonomous multi-step agents.
Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering The paper introduces a test showing that steering vectors designed for competing ethical frameworks behave interchangeably in open-weight models. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Analyzes ethical steering vectors, not code generation or program repair.
Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering The paper introduces a test showing that steering vectors designed for competing ethical frameworks behave interchangeably in open-weight models. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on activation steering evaluation, not model efficiency or quantization.
Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering The paper introduces a test showing that steering vectors designed for competing ethical frameworks behave interchangeably in open-weight models. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Examines text model representation steering, not multimodal perception.
Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering The paper introduces a test showing that steering vectors designed for competing ethical frameworks behave interchangeably in open-weight models. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Deals with internal representation steering, not external document retrieval.
Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems The paper analyzes Best-of-N sampling as an extreme-value competition between tails of correct and incorrect responses to explain reward hacking. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Examines Best-of-N reward sampling, not document retrieval or RAG.
Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems The paper analyzes Best-of-N sampling as an extreme-value competition between tails of correct and incorrect responses to explain reward hacking. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Studies sampling theory rather than model compression or efficient serving.
Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems The paper analyzes Best-of-N sampling as an extreme-value competition between tails of correct and incorrect responses to explain reward hacking. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Theoretical analysis of sampling regimes, not multi-step tool-using agents.
Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems The paper analyzes Best-of-N sampling as an extreme-value competition between tails of correct and incorrect responses to explain reward hacking. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Deals with reward scaling theory, not multimodal data.
Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems The paper analyzes Best-of-N sampling as an extreme-value competition between tails of correct and incorrect responses to explain reward hacking. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Theoretical analysis of reward modeling, not code assistants or software repair.
DANGERLENS: Probing Dangerous Representations Hidden by Safety Alignment The paper introduces an internal probing method to detect whether safety-aligned language models still retain accessible representations of hazardous concepts. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Examines internal safety representation audits, not agent planning or tool use.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on memory representations rather than small or quantized models.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Focuses on multimodal agent memory, not code generation or repair.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 15:06	Not relevant Broad survey of multimodal agents, not industrial defect detection.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on agent memory taxonomy, not deployment safety or attacks.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Provides a conceptual survey framework rather than an evaluation benchmark.
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory The paper provides an analytical framework for how multimodal agents form, represent, and adapt memory over long-horizon tasks. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Analyzes agent cognitive memory, not document retrieval systems.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Evaluates web and retail agents, not software engineering or code LLMs.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Applies strictly to text-based multi-turn agent environments.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Targets general agent task performance rather than safety guardrails or robustness.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Introduces a fine-tuning methodology rather than an evaluation benchmark.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on critique-based fine-tuning rather than model pruning or quantization.
Beyond Corrected Actions: Training LLM Agents on Textual Critiques The paper proposes a fine-tuning method that trains language model agents using textual expert critiques alongside corrected actions. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Focuses on multi-turn action training, not retrieval-augmented generation.
Policy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy Rewards The paper introduces a self-training method that uses runtime policy verification feedback as rewards to teach agents safety compliance. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on policy compliance and preference training, not efficiency or quantization.
Policy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy Rewards The paper introduces a self-training method that uses runtime policy verification feedback as rewards to teach agents safety compliance. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Proposes a training framework rather than an LLM evaluation benchmark.
Evaluating Compressed Memory: Dependence, Utility, and Content Benefit The paper evaluates whether language models genuinely benefit from compressed memory by comparing them against memory-free baseline models. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Studies memory grounding methodology rather than model compression techniques.
Evaluating Compressed Memory: Dependence, Utility, and Content Benefit The paper evaluates whether language models genuinely benefit from compressed memory by comparing them against memory-free baseline models. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Studies passage memory utility in QA, not code generation.
Evaluating Compressed Memory: Dependence, Utility, and Content Benefit The paper evaluates whether language models genuinely benefit from compressed memory by comparing them against memory-free baseline models. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Investigates memory representations in QA/story generation, not agent tool use.
Evaluating Compressed Memory: Dependence, Utility, and Content Benefit The paper evaluates whether language models genuinely benefit from compressed memory by comparing them against memory-free baseline models. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on evaluation methodology rather than deployment safety or jailbreaks.
Evaluating Compressed Memory: Dependence, Utility, and Content Benefit The paper evaluates whether language models genuinely benefit from compressed memory by comparing them against memory-free baseline models. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Focuses exclusively on text passages and memory representations.
DeLoop: Repeated Actions as Self-Supervision for Long-Horizon Agents The paper proposes a self-supervised method that uses repetitive tool calls as negative signals to reduce looping behavior in long-horizon agents. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Introduces a preference optimization training method, not an evaluation benchmark.
EvoWeave: Composing Reusable Agent Experience through Operation Level Graphs The paper introduces an operation-level graph framework that helps interactive agents retrieve and compose reusable fragments from past task experience. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on workflow decomposition and reuse rather than safety or reliability.
EvoWeave: Composing Reusable Agent Experience through Operation Level Graphs The paper introduces an operation-level graph framework that helps interactive agents retrieve and compose reusable fragments from past task experience. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Proposes an agent memory mechanism rather than a standard benchmark.
ixi-sLM: On-Device sLM for AI Call Agent Service The paper presents techniques to adapt and quantize a small language model for on-device deployment in phone call agent services. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on on-device model compression, not safety, guardrails, or hallucination mitigation.
ixi-sLM: On-Device sLM for AI Call Agent Service The paper presents techniques to adapt and quantize a small language model for on-device deployment in phone call agent services. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Engineering report on model deployment, not an evaluation benchmark.
STRATA: Self-Maintaining Structured Experience Graphs for Self-Evolving Language Agents The paper introduces a structured experience graph memory that organizes and maintains agent workflows across hierarchical abstraction levels. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on agent memory retention and organization rather than AI safety.
STRATA: Self-Maintaining Structured Experience Graphs for Self-Evolving Language Agents The paper introduces a structured experience graph memory that organizes and maintains agent workflows across hierarchical abstraction levels. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Proposes an agent memory architecture, not an evaluation framework or benchmark.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Dataset for speech technologies, not multi-step autonomous planning agents.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Presents a multilingual speech dataset rather than an LLM evaluation benchmark.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on linguistic coverage, not model compression or serving efficiency.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Focuses on multilingual speech data, not software engineering or code.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Acoustic speech corpus, not retrieval-augmented generation system.
ELARSpeech: A 30K-Hour Dataset Across 675 Underrepresented Languages for Long-Tail Multilingual Speech Technologies The paper introduces a speech dataset spanning 675 underrepresented languages to support speech recognition and text-to-speech technologies. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Speech recognition dataset, not safety guardrails or jailbreak prevention.
EpiGuard: An Autonomous Infectious-Disease Early Warning Agent Platform The paper presents an autonomous agent platform designed to monitor surveillance data and generate early warnings for infectious disease outbreaks. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Describes an applied autonomous agent platform rather than an LLM benchmark.
EpiGuard: An Autonomous Infectious-Disease Early Warning Agent Platform The paper presents an autonomous agent platform designed to monitor surveillance data and generate early warnings for infectious disease outbreaks. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on disease monitoring agents rather than deployment AI safety risks.
SERO: Learning from Executable Oracle Continuations for Long-Horizon Agentic Reinforcement Learning The paper proposes an agentic reinforcement learning method that uses executable oracle continuations to guide exploration during long-horizon tasks. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Presents an RL training method rather than an evaluation benchmark.
SERO: Learning from Executable Oracle Continuations for Long-Horizon Agentic Reinforcement Learning The paper proposes an agentic reinforcement learning method that uses executable oracle continuations to guide exploration during long-horizon tasks. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on agent policy optimization rather than safety guardrails or alignment.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on linguistic lexical evaluation rather than safety or reliability.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Multi-agent interaction is only a minor experimental subset for coining words.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Studies vocabulary acquisition and generation, not retrieval-augmented generation.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Focuses on language acquisition evaluation, not model compression or efficiency.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Evaluates natural language neologisms, not software code generation.
KoAGE: Do Large Language Models Know, Learn, and Coin Korean Neologisms? The paper introduces a benchmark to evaluate how well language models recognize, infer, and coin newly emerging Korean terms. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Deals strictly with text-based neologisms, not multimodal data.
Scaling LLM Inference via Parallelism: Throughput-Latency Trade-off The paper presents a systematic framework for selecting optimal parallelism strategies to balance latency and throughput during distributed language model serving. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Focuses on inference systems profiling rather than general LLM benchmarking.
Scaling LLM Inference via Parallelism: Throughput-Latency Trade-off The paper presents a systematic framework for selecting optimal parallelism strategies to balance latency and throughput during distributed language model serving. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Covers serving systems engineering, not safety, jailbreaks, or robustness.
Scaling LLM Inference via Parallelism: Throughput-Latency Trade-off The paper presents a systematic framework for selecting optimal parallelism strategies to balance latency and throughput during distributed language model serving. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Addresses distributed serving systems, not multi-step planning or tool use.

7,441–7,500 of 17,313← Previous Next →