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
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents Introduces a framework allowing agents to rewind long-horizon environments to earlier states while retaining learned exploratory knowledge. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Uses established benchmarks to assess agent backtracking rather than proposing new evaluations.
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents Introduces a framework allowing agents to rewind long-horizon environments to earlier states while retaining learned exploratory knowledge. Preprint	Multimodal models via arxiv · 12 Oct, 16:28	Not relevant Focuses on decision-making policies in interactive text/grid environments, not multimodal models.
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents Introduces a framework allowing agents to rewind long-horizon environments to earlier states while retaining learned exploratory knowledge. Preprint	Retrieval-augmented generation (RAG) via arxiv · 12 Oct, 16:28	Not relevant Deals with reinforcement learning and state rollbacks, not document retrieval or grounding.
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents Introduces a framework allowing agents to rewind long-horizon environments to earlier states while retaining learned exploratory knowledge. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Focuses on exploration policies and state rewinding, not model compression or efficient serving.
Travel Back To Move Forward: Retrospective Exploration and Evolving for Long Horizon Agents Introduces a framework allowing agents to rewind long-horizon environments to earlier states while retaining learned exploratory knowledge. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Focuses on task-completion search policies rather than safety guardrails or adversarial robustness.
Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment Proposes an advertising image editing method that identifies and selectively refines visual attributes driving click-through rates. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Proposes an advertising image optimization pipeline, not an LLM or vision evaluation benchmark.
Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment Proposes an advertising image editing method that identifies and selectively refines visual attributes driving click-through rates. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Optimizes marketing conversion metrics rather than model safety, security, or robustness.
Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment Proposes an advertising image editing method that identifies and selectively refines visual attributes driving click-through rates. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on CTR-driven image editing rather than autonomous planning or agent tool use.
Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment Proposes an advertising image editing method that identifies and selectively refines visual attributes driving click-through rates. Preprint	Computer vision for industrial applications via arxiv · 12 Oct, 16:28	Not relevant Deals with e-commerce ad creative editing, not industrial manufacturing or robotic vision.
LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence Management Presents an agent harness that coevolves evidence and conclusions to improve reasoning across 100-million-token document corpora. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Introduces an agent framework for corpus-level reasoning, not a general evaluation benchmark.
LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence Management Presents an agent harness that coevolves evidence and conclusions to improve reasoning across 100-million-token document corpora. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Focuses on information retrieval and reasoning state tracking rather than safety or guardrails.
Attention Sees, MLP Refuses: Tracing the Attention-to-MLP Safety Pathway in LLMs Investigates internal transformer mechanisms, showing attention updates decode harmfulness while MLP layers primarily drive refusal token generation. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Investigates internal transformer layer mechanics, not autonomous multi-step agents or tool use.
Attention Sees, MLP Refuses: Tracing the Attention-to-MLP Safety Pathway in LLMs Investigates internal transformer mechanisms, showing attention updates decode harmfulness while MLP layers primarily drive refusal token generation. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Focuses on mechanistic interpretability of safety mechanisms rather than benchmarking LLM systems.
SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty Presents a difficulty-aware framework that allocates reasoning steps across sections to improve efficiency in large audio-language models. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Focuses on an adaptive inference framework evaluated on standard benchmarks, not evaluation design.
SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty Presents a difficulty-aware framework that allocates reasoning steps across sections to improve efficiency in large audio-language models. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Examines internal token-budget allocation for audio reasoning, not agent planning or tool use.
SWEAR: Section-Wise Efficient Audio Reasoning via Self-Rated Difficulty Presents a difficulty-aware framework that allocates reasoning steps across sections to improve efficiency in large audio-language models. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Addresses inference efficiency and token allocation rather than safety, guardrails, or alignment.
VETR: Execution-Grounded Self-Improving Agent for Symbolic Graphics Reasoning Introduces a recursive self-improvement agent framework that refines symbolic graphics programs using execution traces and runtime state changes. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Focuses on symbolic graphics reasoning and refinement rather than deployment safety or guardrails.
VETR: Execution-Grounded Self-Improving Agent for Symbolic Graphics Reasoning Introduces a recursive self-improvement agent framework that refines symbolic graphics programs using execution traces and runtime state changes. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Evaluates on existing Blender benchmarks rather than introducing a new evaluation suite.
Does Location Matter? A Black-Box Causal Audit Reveals Model-Dependent Audio Grounding in LALM Hallucination Audits audio-language model hallucination mitigations by testing localized acoustic evidence ablation against baseline methods across model architectures. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Studies multimodal audio-language decoding and hallucination, not tool-using or autonomous agent systems.
Does Location Matter? A Black-Box Causal Audit Reveals Model-Dependent Audio Grounding in LALM Hallucination Audits audio-language model hallucination mitigations by testing localized acoustic evidence ablation against baseline methods across model architectures. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Focuses on causal grounding and contrastive decoding, not model compression or efficient serving.
Continually Post-training Small Models via Synthetic Self-distillation Proposes a synthetic self-distillation method using a slightly larger aligner to enable continual post-training in small language models. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Applies standard retention benchmarks to evaluate continual learning rather than proposing new evaluation methods.
Continually Post-training Small Models via Synthetic Self-distillation Proposes a synthetic self-distillation method using a slightly larger aligner to enable continual post-training in small language models. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on continual post-training distillation, not agentic planning, multi-step actions, or tool use.
Continually Post-training Small Models via Synthetic Self-distillation Proposes a synthetic self-distillation method using a slightly larger aligner to enable continual post-training in small language models. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Addresses catastrophic forgetting in small models rather than AI safety, guardrails, or red teaming.
Locate, Judge, Read: Role-Specialized Tool Agents over a Knowledge-Point Index for Open-Domain Multimodal Document QA Designs a modular multi-agent pipeline separating search, verification, and reading for multimodal document question answering. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Evaluates on existing benchmarks rather than proposing novel LLM evaluation methodologies.
Locate, Judge, Read: Role-Specialized Tool Agents over a Knowledge-Point Index for Open-Domain Multimodal Document QA Designs a modular multi-agent pipeline separating search, verification, and reading for multimodal document question answering. Preprint	LLMs for code via arxiv · 12 Oct, 16:28	Not relevant Applies code execution solely as an internal scheduler for document QA, not software engineering.
Locate, Judge, Read: Role-Specialized Tool Agents over a Knowledge-Point Index for Open-Domain Multimodal Document QA Designs a modular multi-agent pipeline separating search, verification, and reading for multimodal document question answering. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Uses a 35B model with 3B active parameters, but efficiency is not the primary research focus.
Locate, Judge, Read: Role-Specialized Tool Agents over a Knowledge-Point Index for Open-Domain Multimodal Document QA Designs a modular multi-agent pipeline separating search, verification, and reading for multimodal document question answering. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Focuses on QA accuracy and retrieval orchestration rather than safety, guardrails, or reliability.
Looking Ahead and Back: Anticipatory Generative Modeling for Road-Network Trajectories Introduces an anticipatory generative modeling approach for road networks to improve realistic next-transition selection in trajectory generation. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Focuses on generative trajectory modeling methods rather than language model evaluation frameworks.
Looking Ahead and Back: Anticipatory Generative Modeling for Road-Network Trajectories Introduces an anticipatory generative modeling approach for road networks to improve realistic next-transition selection in trajectory generation. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Addresses urban trajectory generation rather than LLM safety, guardrails, or alignment risks.
Looking Ahead and Back: Anticipatory Generative Modeling for Road-Network Trajectories Introduces an anticipatory generative modeling approach for road networks to improve realistic next-transition selection in trajectory generation. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Does not address model compression, quantization, distillation, or efficient LLM serving.
Looking Ahead and Back: Anticipatory Generative Modeling for Road-Network Trajectories Introduces an anticipatory generative modeling approach for road networks to improve realistic next-transition selection in trajectory generation. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Investigates trajectory generation on road graphs rather than LLM-based autonomous agents or tool use.
DP-MIND: Differentially private memory-guided Long context Inference Proposes a memory-guided differential privacy framework enabling private inference over million-token contexts in LLM-as-a-service settings. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on private long-context inference and memory mechanisms, not multi-step agent planning or tool use.
DP-MIND: Differentially private memory-guided Long context Inference Proposes a memory-guided differential privacy framework enabling private inference over million-token contexts in LLM-as-a-service settings. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Proposes a privacy-preserving inference framework rather than evaluating models or introducing a general benchmark.
Competence-Aware Orchestration for LLM-Based Multi-Agent Systems Presents a framework that dynamically adjusts role activation and multi-agent collaboration based on the model backbone's query-specific competence. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not evaluate safety, jailbreaks, prompt injection, or alignment defenses.
Competence-Aware Orchestration for LLM-Based Multi-Agent Systems Presents a framework that dynamically adjusts role activation and multi-agent collaboration based on the model backbone's query-specific competence. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Focuses on multi-agent orchestration architecture rather than evaluation methodology or benchmark metrics.
VIGOR: Verifiable Industrial Geometry Optimization and Repair in Scene Programs Introduces an execution-grounded dataset and method to repair geometric and engineering constraint violations in LLM-generated industrial scene programs. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on scene program synthesis and geometric solver optimization rather than interactive agents.
Quality, Not Just Outcome: A Structural Scoring Framework for Selecting Coding-Agent Trajectories for Supervised Fine-Tuning Presents a structural scoring framework to select high-quality coding agent trajectories for supervised fine-tuning based on interaction efficiency. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Focuses on training data selection rather than deployment guardrails or safety red-teaming.
RoleWorld: Knowledge Transfer in World Modeling with Functional Role Abstractions Proposes a symbolic world modeling method using functional role abstractions to transfer interaction dynamics across environments for LLM agents. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Focuses on agent world modeling techniques rather than benchmark design or LLM evaluation.
RoleWorld: Knowledge Transfer in World Modeling with Functional Role Abstractions Proposes a symbolic world modeling method using functional role abstractions to transfer interaction dynamics across environments for LLM agents. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not study AI safety risks, prompt injections, guardrails, or hallucination.
AdaC-GRPO: Towards Mastering Long-Horizon Visual Sudoku Puzzles via Adaptive Curriculum Learning Introduces an adaptive curriculum reinforcement learning method to train multimodal models on long-horizon visual Sudoku reasoning tasks. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not address safety, guardrails, prompt injection, or alignment risks.
AdaC-GRPO: Towards Mastering Long-Horizon Visual Sudoku Puzzles via Adaptive Curriculum Learning Introduces an adaptive curriculum reinforcement learning method to train multimodal models on long-horizon visual Sudoku reasoning tasks. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on post-training multimodal LLMs via RLVR, not autonomous multi-step tool-using agents.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	Retrieval-augmented generation (RAG) via arxiv · 12 Oct, 16:28	Not relevant Does not investigate retrieval-augmented generation or enterprise document retrieval systems.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	LLMs for code via arxiv · 12 Oct, 16:28	Not relevant Focuses on scientific research discovery workflows rather than software engineering or code assistants.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Introduces an autonomous agent framework rather than testing benchmark contamination or judge methods.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Does not cover model compression, distillation, quantization, or efficient deployment.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	Multimodal models via arxiv · 12 Oct, 16:28	Not relevant Does not address multimodal reasoning across images, audio, or video modalities.
JANUS: Enhancing Long-Horizon Autonomous Research with Co-Evolving World Models Proposes JANUS, using a co-evolving world model to predict, verify, and guide exploration paths in autonomous scientific research. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not study AI safety guardrails, jailbreaks, or red-teaming reliability.
Does Learning Recurrent Networks Matter in Partially Observable RL? Investigates whether frozen recurrent networks can provide sufficient memory for partially observable reinforcement learning without end-to-end training. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Studies recurrent architectures in POMDP reinforcement learning, not LLM agents.
Does Learning Recurrent Networks Matter in Partially Observable RL? Investigates whether frozen recurrent networks can provide sufficient memory for partially observable reinforcement learning without end-to-end training. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Investigates memory in POMDPs rather than LLM benchmarking or evaluation frameworks.
Does Learning Recurrent Networks Matter in Partially Observable RL? Investigates whether frozen recurrent networks can provide sufficient memory for partially observable reinforcement learning without end-to-end training. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not examine safety guardrails, red teaming, hallucination, or prompt injection.
Beyond Actions: Learning Future Operation Targets for Vision-Language-Action Models Introduces AFCE, training vision-language-action policies to predict future physical effects rather than raw action chunks directly. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Does not address language model evaluation benchmarks or LLM testing methodologies.
Beyond Actions: Learning Future Operation Targets for Vision-Language-Action Models Introduces AFCE, training vision-language-action policies to predict future physical effects rather than raw action chunks directly. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on robotic low-level policy action control rather than autonomous LLM agents.
Beyond Actions: Learning Future Operation Targets for Vision-Language-Action Models Introduces AFCE, training vision-language-action policies to predict future physical effects rather than raw action chunks directly. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not address LLM safety risks, guardrails, hallucination, or adversarial attacks.
ReSFlow: Rectifying Categorical Flows on Statistical Manifolds for One-Step Generation Introduces ReSFlow, rectifying categorical flows along geodesic trajectories on statistical manifolds to enable one-step discrete generation. Preprint	AI agents via arxiv · 12 Oct, 16:28	Not relevant Focuses on continuous generative flow mathematics, not LLM agents or planning.
ReSFlow: Rectifying Categorical Flows on Statistical Manifolds for One-Step Generation Introduces ReSFlow, rectifying categorical flows along geodesic trajectories on statistical manifolds to enable one-step discrete generation. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:28	Not relevant Does not study LLM evaluation benchmarks or assessment techniques.
ReSFlow: Rectifying Categorical Flows on Statistical Manifolds for One-Step Generation Introduces ReSFlow, rectifying categorical flows along geodesic trajectories on statistical manifolds to enable one-step discrete generation. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:28	Not relevant Does not cover AI alignment, security attacks, guardrails, or safety deployment.
How Reliable are AMR Parser Outputs? Uncertainty Estimation for AMR Parsing Studies uncertainty estimation for abstract meaning representation parsers using top-k beam candidates and prediction rejection ratios. Preprint	Retrieval-augmented generation (RAG) via arxiv · 12 Oct, 16:28	Not relevant Does not involve retrieval-augmented generation or external document indexing.
How Reliable are AMR Parser Outputs? Uncertainty Estimation for AMR Parsing Studies uncertainty estimation for abstract meaning representation parsers using top-k beam candidates and prediction rejection ratios. Preprint	Multimodal models via arxiv · 12 Oct, 16:28	Not relevant Does not investigate multimodal inputs like images, video, or audio.
How Reliable are AMR Parser Outputs? Uncertainty Estimation for AMR Parsing Studies uncertainty estimation for abstract meaning representation parsers using top-k beam candidates and prediction rejection ratios. Preprint	LLMs for code via arxiv · 12 Oct, 16:28	Not relevant Studies semantic parsing graphs, not source code generation or software engineering.
How Reliable are AMR Parser Outputs? Uncertainty Estimation for AMR Parsing Studies uncertainty estimation for abstract meaning representation parsers using top-k beam candidates and prediction rejection ratios. Preprint	Small and efficient models via arxiv · 12 Oct, 16:28	Not relevant Does not address quantization, distillation, or efficient model compression.

17,041–17,100 of 17,313← Previous Next →