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
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	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Targets scientific equation fitting rather than AI safety or guardrails.
Auditing Failure Attribution: Label Predictability and Judge Behavior This paper provides four auditing tests to determine whether failure-attribution judges actually perform multi-step checks or merely rely on spurious surface cues. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Does not address model compression, distillation, or serving efficiency.
Auditing Failure Attribution: Label Predictability and Judge Behavior This paper provides four auditing tests to determine whether failure-attribution judges actually perform multi-step checks or merely rely on spurious surface cues. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on auditing LLM evaluation judges rather than multi-step planning agents.
Auditing Failure Attribution: Label Predictability and Judge Behavior This paper provides four auditing tests to determine whether failure-attribution judges actually perform multi-step checks or merely rely on spurious surface cues. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Analyzes text-based LLM judges rather than multimodal vision-language architectures.
Auditing Failure Attribution: Label Predictability and Judge Behavior This paper provides four auditing tests to determine whether failure-attribution judges actually perform multi-step checks or merely rely on spurious surface cues. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Does not address retrieval-augmented generation or document retrieval.
Auditing Failure Attribution: Label Predictability and Judge Behavior This paper provides four auditing tests to determine whether failure-attribution judges actually perform multi-step checks or merely rely on spurious surface cues. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Examines failure attribution cues in benchmark tasks rather than software generation or repair.
I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation This paper presents a system combining interactive human corrections with iterative policy retraining for long-horizon whole-body mobile robot manipulation. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Addresses robotic manipulation failures rather than LLM safety, guardrails, or alignment.
I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation This paper presents a system combining interactive human corrections with iterative policy retraining for long-horizon whole-body mobile robot manipulation. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on physical robot teleoperation and imitation learning, not LLM agents.
I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation This paper presents a system combining interactive human corrections with iterative policy retraining for long-horizon whole-body mobile robot manipulation. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Proposes an imitation learning system rather than an LLM evaluation benchmark.
I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation This paper presents a system combining interactive human corrections with iterative policy retraining for long-horizon whole-body mobile robot manipulation. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Addresses whole-body mobile manipulation rather than industrial visual inspection or defect detection.
Degrees of Freedom in Multi-Agent Coordination: Between Centralized and Decentralized Regimes This paper introduces a framework that quantifies delegation flexibility to analyze the spectrum between centralized and decentralized multi-agent coordination. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Studies agent coordination trade-offs rather than safety guardrails or alignment risks.
Degrees of Freedom in Multi-Agent Coordination: Between Centralized and Decentralized Regimes This paper introduces a framework that quantifies delegation flexibility to analyze the spectrum between centralized and decentralized multi-agent coordination. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Investigates organizational routing between agents rather than model compression or inference efficiency.
Degrees of Freedom in Multi-Agent Coordination: Between Centralized and Decentralized Regimes This paper introduces a framework that quantifies delegation flexibility to analyze the spectrum between centralized and decentralized multi-agent coordination. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Introduces an architectural coordination framework rather than an LLM evaluation benchmark.
Instructions Are Evidence, Not Objectives: Measuring Outcome Alignment of Agents This paper introduces an approach and benchmark for evaluating whether language model agents can infer unstated true requirements instead of following incomplete instructions. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Addresses instruction completeness and intent discovery rather than safety guardrails or red-teaming.
Gaussian Process Bandits for AI System Optimization with Selectively Verified Feedback This paper proposes Gaussian process bandit algorithms with selective verification audits to optimize AI system configurations under noisy evaluations. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Focuses on statistical GP bandit verification theory rather than standard LLM benchmarks.
Gaussian Process Bandits for AI System Optimization with Selectively Verified Feedback This paper proposes Gaussian process bandit algorithms with selective verification audits to optimize AI system configurations under noisy evaluations. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Proposes a Bayesian optimization method for audit designs, not LLM agent systems.
Gaussian Process Bandits for AI System Optimization with Selectively Verified Feedback This paper proposes Gaussian process bandit algorithms with selective verification audits to optimize AI system configurations under noisy evaluations. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses GP bandit optimization theory rather than neural network compression or serving.
Autonomous 3D Visual-Thermal Inspection for Utility-Scale Solar Farms This paper presents an autonomous system combining LiDAR SLAM, visual imagery, and thermal sensing to inspect under-panel components of solar farms. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Covers hardware defect inspection rather than AI model safety and guardrails.
Autonomous 3D Visual-Thermal Inspection for Utility-Scale Solar Farms This paper presents an autonomous system combining LiDAR SLAM, visual imagery, and thermal sensing to inspect under-panel components of solar farms. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on robotics and sensor fusion rather than language model planning and tool use.
Autonomous 3D Visual-Thermal Inspection for Utility-Scale Solar Farms This paper presents an autonomous system combining LiDAR SLAM, visual imagery, and thermal sensing to inspect under-panel components of solar farms. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Focuses on 3D computer vision and LiDAR-thermal fusion rather than vision-language models.
Autonomous 3D Visual-Thermal Inspection for Utility-Scale Solar Farms This paper presents an autonomous system combining LiDAR SLAM, visual imagery, and thermal sensing to inspect under-panel components of solar farms. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents an industrial inspection system rather than an LLM benchmark.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Optimizes reasoning tokens generally rather than focusing on software engineering or code generation.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on compute efficiency and adaptive thinking rather than AI safety or alignment.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Operates on text-based reasoning models without multimodal inputs.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Does not address document retrieval or knowledge grounding.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents an optimization recipe rather than an evaluation benchmark or judge audit.
Efficiently Learning To Reason or Not to Reason: Root-token Policy Optimization for Adaptive Thinking This paper presents an optimization method that trains models to decide when to invoke reasoning tokens, reducing unnecessary thinking overhead. Peer-reviewed	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on token-level policy optimization for reasoning generation rather than multi-step agents.
Agentic4PDM: Host-Governed Evidence Admissibility for Industrial Fault Diagnosis This paper introduces a framework combining language model agents with deterministic software controls for reliable sensor-based industrial fault diagnosis. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on agent-driven fault diagnosis rather than model distillation or efficient inference.
Agentic4PDM: Host-Governed Evidence Admissibility for Industrial Fault Diagnosis This paper introduces a framework combining language model agents with deterministic software controls for reliable sensor-based industrial fault diagnosis. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Pertains to industrial sensor diagnosis, not LLM guardrails, hallucination, or safety.
Agentic4PDM: Host-Governed Evidence Admissibility for Industrial Fault Diagnosis This paper introduces a framework combining language model agents with deterministic software controls for reliable sensor-based industrial fault diagnosis. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents a diagnostic framework rather than a general LLM benchmark or evaluation methodology.
Posterior-Averaged Bellman Flow Matching for Distributional Critics This paper proposes a flow-matching method that averages Bellman target directions under a posterior distribution to improve reinforcement learning critics. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses temporal-difference flow matching rather than model compression or efficient serving.
Posterior-Averaged Bellman Flow Matching for Distributional Critics This paper proposes a flow-matching method that averages Bellman target directions under a posterior distribution to improve reinforcement learning critics. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Proposes an algorithmic RL training method rather than an LLM evaluation benchmark.
Posterior-Averaged Bellman Flow Matching for Distributional Critics This paper proposes a flow-matching method that averages Bellman target directions under a posterior distribution to improve reinforcement learning critics. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Pure RL theory without a focus on LLM safety, guardrails, or alignment.
Posterior-Averaged Bellman Flow Matching for Distributional Critics This paper proposes a flow-matching method that averages Bellman target directions under a posterior distribution to improve reinforcement learning critics. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on theoretical reinforcement learning distributional critics, not LLM agents.
GIFT-Eval-CTX: A Benchmark For Contextual Forecasting Model Evaluation This paper introduces a benchmark to evaluate whether forecasting models can incorporate forward-looking contextual information into accurate time-series predictions. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Forecasting benchmark paper; agentic pipelines are used merely to build annotations.
GIFT-Eval-CTX: A Benchmark For Contextual Forecasting Model Evaluation This paper introduces a benchmark to evaluate whether forecasting models can incorporate forward-looking contextual information into accurate time-series predictions. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Evaluates time series forecasting rather than efficient model architectures or inference.
GIFT-Eval-CTX: A Benchmark For Contextual Forecasting Model Evaluation This paper introduces a benchmark to evaluate whether forecasting models can incorporate forward-looking contextual information into accurate time-series predictions. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on numerical time series forecasting accuracy rather than AI safety or guardrails.
Learning from fracture: Teaching language models to design through physical experience This paper introduces a framework that enables language models to optimize physical structural designs through simulated fracture experience and hypothesis revision. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on structural fracture optimization, not AI safety or deployment reliability.
Learning from fracture: Teaching language models to design through physical experience This paper introduces a framework that enables language models to optimize physical structural designs through simulated fracture experience and hypothesis revision. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents a scientific design framework rather than a general LLM evaluation benchmark.
Learning from fracture: Teaching language models to design through physical experience This paper introduces a framework that enables language models to optimize physical structural designs through simulated fracture experience and hypothesis revision. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Does not address model compression, quantization, or efficient inference techniques.
Empowering Edge Agents: A Systematic Analysis of Edge-Cloud Collaboration in LLM Agent Execution This paper presents a benchmark to evaluate performance, resource usage, and privacy trade-offs when language model agents collaborate across edge and cloud environments. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Examines privacy exposure alongside resource trade-offs, but its main focus is agent benchmarking and orchestration.
Same Attack, Different Judge: A First Controlled Adversarial Comparison of Jev, LLM Judges, and Purpose-Built Safety Classifiers This paper conducts an adversarial evaluation comparing a probability-based safety judge against large language model judges and safety classifiers across multiple attack types. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Evaluates safety classification tools rather than multi-step agent planning or tool use.
Checking the Checkers: A Synthesis of Verifier Noise and Robustness in RLVR This survey analyzes how verifier noise affects reinforcement learning with verifiable rewards across 58 studies and summarizes practical mitigation techniques. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Surveys verifier noise in RLVR rather than building or deploying LLM agents.
Checking the Checkers: A Synthesis of Verifier Noise and Robustness in RLVR This survey analyzes how verifier noise affects reinforcement learning with verifiable rewards across 58 studies and summarizes practical mitigation techniques. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Addresses RL reward verifier errors rather than model compression, quantization, or serving efficiency.
Checking the Checkers: A Synthesis of Verifier Noise and Robustness in RLVR This survey analyzes how verifier noise affects reinforcement learning with verifiable rewards across 58 studies and summarizes practical mitigation techniques. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant Focuses on reasoning verifiers in LLMs, not multimodal architectures.
Checking the Checkers: A Synthesis of Verifier Noise and Robustness in RLVR This survey analyzes how verifier noise affects reinforcement learning with verifiable rewards across 58 studies and summarizes practical mitigation techniques. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Focuses on RLVR reward verification, not retrieval-augmented generation.
Checking the Checkers: A Synthesis of Verifier Noise and Robustness in RLVR This survey analyzes how verifier noise affects reinforcement learning with verifiable rewards across 58 studies and summarizes practical mitigation techniques. Preprint	LLMs for code via arxiv · 15 Oct, 06:17	Not relevant Considers unit tests only as an example of reward checkers, not LLMs for code.
QK-Edit: Revisiting Attention-based Injection in MM-DiT for Image and Video Editing Introduces QK-Edit, a training-free query-key attention manipulation method for text-guided image and video editing in multimodal diffusion transformers. Peer-reviewed	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Does not address AI safety, jailbreaks, or deployment guardrails.
QK-Edit: Revisiting Attention-based Injection in MM-DiT for Image and Video Editing Introduces QK-Edit, a training-free query-key attention manipulation method for text-guided image and video editing in multimodal diffusion transformers. Peer-reviewed	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Presents an image/video editing method rather than an evaluation benchmark or methodology.
QK-Edit: Revisiting Attention-based Injection in MM-DiT for Image and Video Editing Introduces QK-Edit, a training-free query-key attention manipulation method for text-guided image and video editing in multimodal diffusion transformers. Peer-reviewed	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on diffusion-based image and video editing, not language model agents.
QK-Edit: Revisiting Attention-based Injection in MM-DiT for Image and Video Editing Introduces QK-Edit, a training-free query-key attention manipulation method for text-guided image and video editing in multimodal diffusion transformers. Peer-reviewed	Computer vision for industrial applications via arxiv · 15 Oct, 06:17	Not relevant Addresses text-guided media editing rather than industrial inspection or manufacturing defect detection.
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Evaluates code editing performance but does not contribute novel general LLM benchmarks or evaluation methods.
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:17	Not relevant Does not address external document retrieval or grounding architectures.
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Focuses on latency and code editing accuracy, not safety, guardrails, or alignment.
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Preprint	Multimodal models via arxiv · 15 Oct, 06:17	Not relevant The framework operates strictly on source code text, not multimodal inputs.
Diff Diffusion: Predicting Edit Paths with Diffusion Models for Efficient Code Editing Proposes Diff Diffusion, a diffusion-based model that predicts localized code edit sequences rather than regenerating entire source files. Preprint	AI agents via arxiv · 15 Oct, 06:17	Not relevant Focuses on diffusion models for code editing rather than multi-step agent planning or tool use.
Agentic-MatriBO: Capability-Compiled Agentic Bayesian Optimization for Materials and Chemical Design Introduces Agentic-MatriBO, a modular Bayesian optimization framework using LLM agents and deterministic compilation for material and chemical experiment design. Preprint	Small and efficient models via arxiv · 15 Oct, 06:17	Not relevant Focuses on Bayesian optimization for chemistry, not model compression or serving.
Agentic-MatriBO: Capability-Compiled Agentic Bayesian Optimization for Materials and Chemical Design Introduces Agentic-MatriBO, a modular Bayesian optimization framework using LLM agents and deterministic compilation for material and chemical experiment design. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:17	Not relevant Proposes an optimization framework, not a general evaluation benchmark.
Agentic-MatriBO: Capability-Compiled Agentic Bayesian Optimization for Materials and Chemical Design Introduces Agentic-MatriBO, a modular Bayesian optimization framework using LLM agents and deterministic compilation for material and chemical experiment design. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Deals with materials design optimization, not deployment safety or jailbreaks.
A Tutorial on Quantum Amplitude Embedding with Data Re-Uploading Provides a tutorial analyzing theoretical expressiveness gains and parameter tradeoffs when re-uploading data in quantum amplitude embedding architectures. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:17	Not relevant Covers quantum circuit encoding, not AI safety or reliability.

13,621–13,680 of 17,313← Previous Next →