arXiv cs.AI
JobBench: Aligning Agent Work With Human Will
The paper introduces JobBench, a new benchmark for evaluating AI agents based on human needs rather than economic…
Hub / Ai Research
ai-research · WeSearch
Ai Research news.
Page 7 of Ai Research headlines on WeSearch — deduped and updated continuously from 10+ editorial sources.
arXiv cs.AI
OmniToM: Benchmarking Theory of Mind in LLMs via Explicit Belief Modeling
The paper introduces OmniToM, a benchmark designed to evaluate the Theory of Mind capabilities in large language…
arXiv cs.AI
Anchor: Mitigating Artifact Drift in Agent Benchmark Generation
The paper introduces Anchor, a task-generation pipeline designed to address artifact drift in AI agent benchmark…
arXiv cs.AI
Experiments in Agentic AI for Science
The paper discusses two innovative frameworks for creating autonomous AI systems to enhance scientific workflows.…
arXiv cs.AI
Your Agents Are Aging Too: Agent Lifespan Engineering for Deployed Systems
The paper discusses the aging of AI agents deployed in operational systems and introduces a new benchmark called…
arXiv cs.AI
Constraint acquisition needs better benchmarks
The paper discusses the need for improved benchmarks in Constraint Acquisition (CA) research. Current benchmarks are…
arXiv cs.AI
Personalizing Embodied Multimodal Large Language Model Agents over Long-term User Interactions
The paper presents POLAR, a framework designed for personalizing embodied multimodal large language model agents…
arXiv cs.AI
Is Agent Memory a Database? Rethinking Data Foundations for Long-Term AI Agent Memory
The paper discusses the need for persistent memory in long-running AI agents. It critiques current memory systems and…
arXiv cs.AI
Can LLMs Introspect? A Reality Check
The paper titled 'Can LLMs Introspect? A Reality Check' questions the ability of large language models (LLMs) to…
arXiv cs.AI
BrickAnything: Geometry-Conditioned Buildable Brick Generation with Structure-Aware Tokenization
The article presents BrickAnything, a new framework for generating buildable brick structures from 3D shapes. This…
arXiv cs.AI
What Gets Cited: Competitive GEO in AI Answer Engines
The paper titled 'What Gets Cited: Competitive GEO in AI Answer Engines' explores how AI answer engines cite sources.…
arXiv cs.AI
Credit Assignment with Resets in Language Model Reasoning
The article discusses advancements in credit assignment methods for language model reasoning in reinforcement…
arXiv cs.AI
ATWL: A Formal Language for Representing, Comparing, and Reusing Visual Analytics Workflows
The paper introduces the Artifact-Transform Workflow Language (ATWL), a formal language designed for visual analytics…
arXiv cs.AI
A Signal-Language Foundation Model for Broad-Spectrum Cardiovascular Assessment from Routine Electrocardiography
A new AI model called ECGCLIP has been developed to enhance cardiovascular assessment from routine…
arXiv cs.AI
Security of OpenClaw Agents: Fundamentals, Attacks, and Countermeasures
The paper discusses the security challenges associated with OpenClaw agents, a new class of autonomous systems. It…
arXiv cs.AI
CODESKILL: Learning Self-Evolving Skills for Coding Agents
CODESKILL is a proposed framework aimed at enhancing coding agents' abilities through self-evolving skills. It…
arXiv cs.AI
Towards end-to-end LLM-based censoring-aware survival analysis
The article discusses a new framework called LLMSurvival that enables censoring-aware survival analysis using large…
arXiv cs.AI
Second Guess: Detecting Uncertainty Through Abstention and Answer Stability in Small Language Models
The paper titled 'Second Guess' introduces a method for detecting uncertainty in small language models through a…
arXiv cs.AI
Context-CoT: Enhancing Context Learning via High-Quality Reasoning Synthesis
The paper titled 'Context-CoT: Enhancing Context Learning via High-Quality Reasoning Synthesis' addresses the…
arXiv cs.AI
AI Cartography: Mapping the Latent Landscape of AI Benchmark Ecosystems
The paper titled 'AI Cartography: Mapping the Latent Landscape of AI Benchmark Ecosystems' introduces a framework for…
arXiv cs.AI
Whose Alignment? Comparing LLM Process Alignment Across Diverse Organizational Decision Contexts
The paper discusses the alignment of AI systems with organizational decision-making, emphasizing the complexity of…
arXiv cs.AI
LipoAgent: Coordinating Fine-Tuned LLM Agents for Safer Lipid Design
LipoAgent is a new framework designed to enhance the safety and efficiency of lipid nanoparticles for nucleic acid…
arXiv cs.AI
FrontierOR: Benchmarking LLMs' Capacity for Efficient Algorithm Design in Large-Scale Optimization
The paper introduces FrontierOR, a benchmark designed to evaluate the capacity of large language models (LLMs) in…
arXiv cs.AI
Meta-Agent: From Task Descriptions to Verified Multi-Agent Systems
The paper introduces Meta-Agent, a framework designed to improve the reliability of multi-agent systems. It automates…
arXiv cs.AI
Boosting Inference with Guided Reasoning: Stochastic Exploration for Recursive Models
The paper discusses a novel approach to enhancing inference in recursive neural networks through guided reasoning and…
arXiv cs.AI
DarkForest: Less Talk, Higher Accuracy for Multi-Agent LLMs
The paper presents DarkForest, a new framework aimed at improving the accuracy of multi-agent large language models…
arXiv cs.AI
SpecAlign: A Semantic Alignment Framework for SystemVerilog Assertion Generation
The paper introduces SpecAlign, a framework designed to enhance the semantic alignment of SystemVerilog Assertions…
arXiv cs.AI
SimuWoB: Simulating Real-World Mobile Apps for Fast and Faithful GUI Agent Benchmarking
The paper introduces SimuWoB, a synthetic benchmark designed for evaluating mobile GUI agents. It addresses the…
arXiv cs.AI
Representation Without Control: Testing the Realization Effect in Language Models
The paper titled 'Representation Without Control: Testing the Realization Effect in Language Models' explores the…
arXiv cs.AI
Beyond the Frontier: Stochastic Backtracking for Efficient Test-Time Scaling
The paper introduces a method called stochastic backtracking for improving test-time scaling in language models. This…
arXiv cs.AI
Trust but Verify: Prover-Verifier Deliberation for Selective LLM Prediction
The paper introduces a new protocol called prover-verifier deliberation (PVD) for improving the reliability of…
arXiv cs.AI
RECTOR: Priority-Aware Rule-Based Reranking for Compliance-Aware Autonomous Driving Trajectory Selection
The paper presents RECTOR, a rule-based reranking system designed for autonomous driving trajectory selection. It…
arXiv cs.AI
Evolutionary Enhanced Multi-Agent Reinforcement Learning for Cooperative Air Combat
The paper presents a new framework for multi-agent reinforcement learning in cooperative air combat scenarios. It…
arXiv cs.AI
AION: Next-Generation Tasks and Practical Harness for Time Series
The paper titled 'AION: Next-Generation Tasks and Practical Harness for Time Series' presents a new framework for time…
arXiv cs.AI
Privacy-Preserving Local Language Models for Longitudinal Data Retrieval in Chronic Dermatologic Disease: Implementation in Pemphigus Patients
A study evaluated the use of a privacy-preserving small language model (SLM) for retrieving clinical data in pemphigus…
arXiv cs.AI
NeurIPS: Neuro-anatomical Inductive Priors for Sphere-based Brain Decoding
The paper presents NeurIPS, a framework designed to enhance surface-based brain decoding by utilizing neuro-anatomical…
arXiv cs.AI
Mitigating Object Hallucinations in Vision-Language Models through Region-Aware Attention Recalibration
A new paper addresses the issue of object hallucination in Large Vision-Language Models (LVLMs). The authors propose a…
arXiv cs.AI
Towards Multi-Turn Dialog Systems for Industrial Asset Operations and Maintenance
The paper presents a multi-turn dialog system tailored for industrial asset operations and maintenance. This system…
arXiv cs.AI
Energy Shields for Fairness
The paper titled 'Energy Shields for Fairness' introduces a novel approach to ensuring runtime fairness in…
arXiv cs.AI
Noise-Robust Financial Numerical Entity Attribute Tagging
The paper introduces a new method called NORA for improving the understanding of financial numerical entities in…
arXiv cs.AI
ProActor: Timing-Aware Reinforcement Learning for Proactive Task Scheduling Agents
The paper introduces ProActor, a framework for proactive task scheduling using timing-aware reinforcement learning. It…
arXiv cs.AI
TaBIIC2: Interactive Building of Ontological Taxonomies using Weighted Self-Organizing Maps
The paper presents TaBIIC2, a tool designed for the interactive construction of ontological taxonomies using weighted…
arXiv cs.AI
Inverting the Shield: Systematically Generating Safety Tests from Policy Specifications
A new framework called POLARIS has been introduced to enhance safety testing for Large Language Models (LLMs). This…
arXiv cs.AI
Clustering as Reasoning: A $k$-Means Interpretation of Chain-of-Thought Graph Learning
The paper presents a novel approach to Chain-of-Thought (CoT) graph learning by interpreting it through the lens of…
arXiv cs.AI
Solving Combinatorial Counting Problems with Weighted First-Order Model Counting
The paper presents a new approach to solving combinatorial counting problems using a language called Cofola. This…
arXiv cs.AI
Geo-Expert: Towards Expert-Level Geological Reasoning via Parameter-Efficient Fine-Tuning
The paper introduces Geo-Expert, a series of parameter-efficient geological language models designed to improve…
arXiv cs.AI
Test-Time Deep Thinking to Explore Implicit Rules
A new framework called Test-Time Exploration (TTExplore) aims to improve the performance of intelligent agents in…
arXiv cs.AI
Agent Manufacturing: Foundation-Model Agents as First-Class Industrial Entities
The article discusses a new paradigm in manufacturing called Agent Manufacturing, which focuses on the role of…
arXiv cs.AI
CoRe-Code: Collaborative Reinforcement Learning for Code Generation
The paper introduces CoRe-Code, a framework for collaborative reinforcement learning aimed at improving code…
arXiv cs.AI
PANDO: Efficient Multimodal AI Agents via Online Skill Distillation
The paper introduces PANDO, a framework designed to enhance the efficiency of multimodal AI agents through online…