AI Agents

Deepmind's 'AI Agent Traps' Paper Maps How Hackers Could Weaponize AI Agents Against Users

Claude, OpenClaw and the new reality: AI agents are here — and so is the chaos

Agentic AI in Beauty: How ChatGPT Is Reshaping Discovery, Trust, and Conversion

[2505.17592] AstroMLab 4: Benchmark-Topping Performance in Astronomy Q&A with a 70B-Parameter Domain-Specialized Reasoning Model

AstroMLab 4 introduces a 70B-parameter AI model specialized for astronomy, achieving benchmark-topping performance in Q&A tasks, surpassi...

[2502.00835] CAIMAN: Causal Action Influence Detection for Sample-efficient Loco-manipulation

The paper introduces CAIMAN, a reinforcement learning framework designed to enhance legged robots' capabilities in non-prehensile loco-ma...

[2408.07110] Physics-informed graph neural networks for flow field estimation in carotid arteries

This article presents a novel approach using physics-informed graph neural networks to estimate hemodynamic flow fields in carotid arteri...

[2602.08227] Investigating Writing Professionals' Relationships with Generative AI: How Combined Perceptions of Rivalry and Collaboration Shape Work Practices and Outcomes

This study explores how writing professionals perceive their relationships with generative AI, highlighting the balance between rivalry a...

[2602.04587] VILLAIN at AVerImaTeC: Verifying Image-Text Claims via Multi-Agent Collaboration

The paper presents VILLAIN, a multimodal fact-checking system that verifies image-text claims through collaborative agents, achieving top...

[2602.02437] UniReason 1.0: A Unified Reasoning Framework for World Knowledge Aligned Image Generation and Editing

UniReason 1.0 presents a unified framework for image generation and editing, integrating textual reasoning and visual refinement to enhan...

[2602.12162] Amortized Molecular Optimization via Group Relative Policy Optimization

The paper presents GRXForm, a novel approach for molecular optimization using Group Relative Policy Optimization, addressing the limitati...

[2601.10161] AWED-FiNER: Agents, Web applications, and Expert Detectors for Fine-grained Named Entity Recognition across 36 Languages for 6.6 Billion Speakers

AWED-FiNER introduces an innovative tool for Fine-grained Named Entity Recognition (FgNER) across 36 languages, enhancing NLP capabilitie...

[2601.20198] DeRaDiff: Denoising Time Realignment of Diffusion Models

The paper presents DeRaDiff, a novel method for denoising time realignment in diffusion models, enabling efficient adjustment of regulari...

[2601.11924] Communication-Corruption Coupling and Verification in Cooperative Multi-Objective Bandits

This paper explores cooperative multi-objective bandits under adversarial corruption, presenting a communication-corruption coupling that...

[2601.01703] Beyond Homophily: Community Search on Heterophilic Graphs

This paper presents Adaptive Community Search (AdaptCS), a novel framework designed to improve community search in heterophilic graphs, o...

[2512.19223] Phase-space entropy at acquisition reflects downstream learnability

The paper explores how phase-space entropy at the acquisition stage can predict the learnability of downstream models, offering a new met...

[2512.08217] Correction of Decoupled Weight Decay

This article discusses the correction of decoupled weight decay in machine learning, challenging the conventional assumption that it shou...

[2512.07805] Group Representational Position Encoding

The paper introduces GRAPE (Group Representational Position Encoding), a framework for positional encoding that integrates multiplicative...

[2512.04954] Amortized Inference of Multi-Modal Posteriors using Likelihood-Weighted Normalizing Flows

This paper introduces a novel technique for amortized posterior estimation using Normalizing Flows, enhancing inference in high-dimension...

[2510.13887] Incomplete Multi-view Clustering via Hierarchical Semantic Alignment and Cooperative Completion

This paper presents a novel framework for incomplete multi-view clustering using Hierarchical Semantic Alignment and Cooperative Completi...

[2512.04388] Learning to Orchestrate Agents in Natural Language with the Conductor

The paper introduces the Conductor model, which utilizes reinforcement learning to optimize coordination strategies among large language ...

[2511.18945] MIST: Mutual Information Estimation Via Supervised Training

The paper presents MIST, a novel approach for estimating mutual information using a neural network trained on a large dataset of syntheti...

[2510.06170] Smartphone-based iris recognition through high-quality visible-spectrum iris image capture.V2

This paper presents a smartphone-based iris recognition system using visible-spectrum imaging, demonstrating high accuracy through a cust...

[2511.02872] FATE: A Formal Benchmark Series for Frontier Algebra of Multiple Difficulty Levels

The paper introduces FATE, a benchmark series for formal algebra, designed to assess large language models' capabilities in advanced math...