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
Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents The paper presents a runtime system that structures long-horizon agent memory around reusable progress objects rather than raw interaction history. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Evaluates web and OS agents rather than software development or coding.
Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents The paper presents a runtime system that structures long-horizon agent memory around reusable progress objects rather than raw interaction history. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Introduces an agent context runtime rather than an evaluation benchmark.
Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents The paper presents a runtime system that structures long-horizon agent memory around reusable progress objects rather than raw interaction history. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Focuses on agent history management, not document retrieval systems or external knowledge databases.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Paper examines general LLM ensemble outputs, not software engineering or code generation.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on static LLM ensemble output aggregation, not autonomous agent planning or multi-step execution.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Deals with text LLM aggregation without multimodal models or inputs.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Studies ensemble aggregation metrics, not model compression or efficient serving.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Analyzes ensemble aggregation metrics rather than retrieval systems, document ranking, or knowledge grounding mechanisms.
Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation The paper analyzes how language model aggregation gains divide into candidate additions and context effects under different baseline comparison setups. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on benchmark aggregation math rather than safety, red-teaming, or guardrails.
Bytes to Threats: Byte-Sequence Transformer For Malware Detection The paper presents a byte-level transformer model designed to detect malware directly from raw executable byte sequences. Peer-reviewed	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents a malware classifier rather than an LLM evaluation framework.
Bytes to Threats: Byte-Sequence Transformer For Malware Detection The paper presents a byte-level transformer model designed to detect malware directly from raw executable byte sequences. Peer-reviewed	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on a malware classification model, not multi-step autonomous agents.
SEAR: Self-Evolving Agentic Robotics The paper presents a framework and benchmark to evaluate whether self-evolving robotic agents genuinely learn reusable skills across tasks. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Studies agent learning and iteration curves, not small model compression.
SEAR: Self-Evolving Agentic Robotics The paper presents a framework and benchmark to evaluate whether self-evolving robotic agents genuinely learn reusable skills across tasks. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Focuses on robotic agent evolution, not retrieval-augmented generation.
SEAR: Self-Evolving Agentic Robotics The paper presents a framework and benchmark to evaluate whether self-evolving robotic agents genuinely learn reusable skills across tasks. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Focuses on robotic agent learning systems, not multimodal model architectures.
SEAR: Self-Evolving Agentic Robotics The paper presents a framework and benchmark to evaluate whether self-evolving robotic agents genuinely learn reusable skills across tasks. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Evaluates learning efficiency in agents, not deployment safety guardrails.
SEAR: Self-Evolving Agentic Robotics The paper presents a framework and benchmark to evaluate whether self-evolving robotic agents genuinely learn reusable skills across tasks. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Addresses general robotic manipulation, not industrial defect detection.
Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning The paper introduces an adaptive diversity control framework that optimizes cooperative multi-agent reinforcement learning teams using episodic return feedback. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on MARL diversity tuning, not AI safety or deployment guardrails.
Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning The paper introduces an adaptive diversity control framework that optimizes cooperative multi-agent reinforcement learning teams using episodic return feedback. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on traditional multi-agent reinforcement learning control, not language model agents.
Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning The paper introduces an adaptive diversity control framework that optimizes cooperative multi-agent reinforcement learning teams using episodic return feedback. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents a control framework for RL, not an LLM benchmark.
Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning The paper introduces an adaptive diversity control framework that optimizes cooperative multi-agent reinforcement learning teams using episodic return feedback. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Addresses multi-agent RL control, not industrial vision defect detection.
Self-Evolving Agent Harness The paper formalizes a framework and theoretical generalization bounds for evolving an agent's software harness without updating model weights. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Focuses on agent harness evolution, not retrieval-augmented generation systems.
Self-Evolving Agent Harness The paper formalizes a framework and theoretical generalization bounds for evolving an agent's software harness without updating model weights. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on agent scaffolding updates, not model quantization or compact architectures.
Self-Evolving Agent Harness The paper formalizes a framework and theoretical generalization bounds for evolving an agent's software harness without updating model weights. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Focuses on agent software harness optimization, not software engineering models.
Self-Evolving Agent Harness The paper formalizes a framework and theoretical generalization bounds for evolving an agent's software harness without updating model weights. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Addresses text-based software scaffolding, not vision or multimodal models.
Self-Evolving Agent Harness The paper formalizes a framework and theoretical generalization bounds for evolving an agent's software harness without updating model weights. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Analyzes harness optimization theory rather than proposing an LLM evaluation benchmark.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Studies machine translation fine-tuning, not autonomous tool-use agents.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Investigates back-translation data, not retrieval-augmented generation systems.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Addresses natural language machine translation, not code generation or repair.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on low-resource translation training, not benchmark design or LLM evaluation.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Deals strictly with text translation, not multimodal models.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Studies translation accuracy, not safety, hallucinations, or deployment guardrails.
Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs The paper examines back-translation using monolingual data as an alternative to continued pretraining for low-resource translation with language models. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Investigates training data for translation, not model pruning or distillation.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Studies decoding temperature calibration, not retrieval-augmented generation systems.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on temperature sampling during text generation, not autonomous multi-step agents.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Proposes scenario-based temperature tuning, not model quantization or compact architectures.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Adapts generation temperature, not guardrails or jailbreak defenses.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Addresses text generation decoding, not multimodal vision-language models.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on a decoding generation method rather than evaluation frameworks.
Read the Room Before Your Model Generates The paper introduces a decoding framework that dynamically selects sampling temperature based on geometric representations of conversational scenarios. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Studies decoding temperature adaptation, not software engineering or code models.
RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks The paper presents an interactive simulation toolkit that enables humans and code agents to iteratively calibrate robot manipulation physics. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Addresses simulation contact dynamics, not AI safety or red-teaming.
RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks The paper presents an interactive simulation toolkit that enables humans and code agents to iteratively calibrate robot manipulation physics. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on physics modeling, not small model distillation or inference efficiency.
RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks The paper presents an interactive simulation toolkit that enables humans and code agents to iteratively calibrate robot manipulation physics. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents simulation software, not LLM evaluation methods or benchmarks.
RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks The paper presents an interactive simulation toolkit that enables humans and code agents to iteratively calibrate robot manipulation physics. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Robotic physics simulation toolkit, not industrial visual quality control.
RoboSimStudio: An Interactive Toolkit for Rapid Real2Sim of Robot Manipulation Tasks The paper presents an interactive simulation toolkit that enables humans and code agents to iteratively calibrate robot manipulation physics. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on an interactive simulation toolkit, not autonomous multi-step LLM agents.
Fast Rates for Offline Q-Learning with KL Regularization The paper analyzes sample complexity rates for offline Q-learning algorithms using Kullback-Leibler regularization. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Theoretical reinforcement learning paper, not deployment safety or guardrails.
Fast Rates for Offline Q-Learning with KL Regularization The paper analyzes sample complexity rates for offline Q-learning algorithms using Kullback-Leibler regularization. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Theoretical offline RL sample complexity analysis, not practical LLM agents.
Fast Rates for Offline Q-Learning with KL Regularization The paper analyzes sample complexity rates for offline Q-learning algorithms using Kullback-Leibler regularization. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Mathematical proof of Q-learning bounds, not LLM evaluation or benchmarks.
Fast Rates for Offline Q-Learning with KL Regularization The paper analyzes sample complexity rates for offline Q-learning algorithms using Kullback-Leibler regularization. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on reinforcement learning convergence rates, not model efficiency or distillation.
Who Improves the Robot? Toward Physical Recursive Self-Improvement The paper analyzes physical recursive self-improvement in robotics, proposing a multi-level evaluation framework and reviewing existing adaptation methods. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Discusses physical self-improvement frameworks, not model quantization or pruning.
Who Improves the Robot? Toward Physical Recursive Self-Improvement The paper analyzes physical recursive self-improvement in robotics, proposing a multi-level evaluation framework and reviewing existing adaptation methods. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Surveys physical recursive self-improvement in robotics, not software LLM agent workflows.
Who Improves the Robot? Toward Physical Recursive Self-Improvement The paper analyzes physical recursive self-improvement in robotics, proposing a multi-level evaluation framework and reviewing existing adaptation methods. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Conceptual framework survey, not an empirical benchmark for language models.
Who Improves the Robot? Toward Physical Recursive Self-Improvement The paper analyzes physical recursive self-improvement in robotics, proposing a multi-level evaluation framework and reviewing existing adaptation methods. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Focuses on robotic self-improvement, not industrial visual defect detection.
LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models The paper introduces a benchmark to measure vision-language-action model robustness against hardware, optical, and physical robot setup variations. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Evaluates robotic hardware variations, not LLM agent planning or tool execution.
LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models The paper introduces a benchmark to measure vision-language-action model robustness against hardware, optical, and physical robot setup variations. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Benchmarking robotic arm perturbations, not industrial manufacturing defect inspection.
MFG: Multi-turn Feedback Guided Molecular Optimization The paper presents a reinforcement learning framework that trains language models to optimize molecules across multiple turns using iterative feedback. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on drug molecule generation, not AI deployment safety or robustness.
MFG: Multi-turn Feedback Guided Molecular Optimization The paper presents a reinforcement learning framework that trains language models to optimize molecules across multiple turns using iterative feedback. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Applies RL to chemistry tasks rather than general LLM evaluation benchmarks.
MFG: Multi-turn Feedback Guided Molecular Optimization The paper presents a reinforcement learning framework that trains language models to optimize molecules across multiple turns using iterative feedback. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses drug design RL, not model compression or efficient serving.
MFG: Multi-turn Feedback Guided Molecular Optimization The paper presents a reinforcement learning framework that trains language models to optimize molecules across multiple turns using iterative feedback. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on molecular optimization via RL, not general tool-use agents.
Two Clocks: Why Larger Rollout Batches Learn Slower but Fail on Schedule in Reinforcement Learning with Verifiable Rewards The paper develops a framework explaining why large rollout batches in reinforcement learning reduce sample efficiency despite maintaining drift thresholds. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on reinforcement learning batch scaling dynamics, not model compression or serving.
Two Clocks: Why Larger Rollout Batches Learn Slower but Fail on Schedule in Reinforcement Learning with Verifiable Rewards The paper develops a framework explaining why large rollout batches in reinforcement learning reduce sample efficiency despite maintaining drift thresholds. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Studies RL optimization dynamics, not autonomous LLM tool-use agents.

10,801–10,860 of 13,754← Previous Next →