Ai Agents
Deepmind's 'AI Agent Traps' Paper Maps How Hackers Could Weaponize AI Agents Against Users
AI Tools & Products
·
7 min
·
AI Agents
Autonomous agents, tool use, and agentic systems
Top This Week
All Content
![[2510.09658] Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
Llms
[2510.09658] Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models
This paper presents Gradient-Sign Masking, a method for transferring task vectors across pre-trained models without additional fine-tunin...
arXiv - Machine Learning
·
4 min
·
Machine Learning
[2510.08570] Who Said Neural Networks Aren't Linear?
This paper explores the linearity of neural networks by introducing a framework that identifies non-standard vector spaces where neural n...
arXiv - Machine Learning
·
4 min
·
Machine Learning
[2507.11551] Landmark Detection for Medical Images using a General-purpose Segmentation Model
The paper presents a novel approach to anatomical landmark detection in medical images by combining YOLO and SAM models, enhancing segmen...
arXiv - AI
·
4 min
·
Machine Learning
[2509.21154] GRPO is Secretly a Process Reward Model
The paper presents a theoretical proof that the GRPO algorithm, typically viewed as an outcome reward model, can be interpreted as a proc...
arXiv - Machine Learning
·
4 min
·
Nlp
[2503.18980] CAE: Repurposing the Critic as an Explorer in Deep Reinforcement Learning
The paper introduces CAE, a novel approach in deep reinforcement learning that repurposes value networks to enhance exploration efficienc...
arXiv - Machine Learning
·
3 min
·
Llms
[2503.16021] Imitating AI agents increase diversity in homogeneous information environments but can reduce it in heterogeneous ones
This article explores how AI agents imitating human content affect information diversity, revealing context-dependent outcomes in homogen...
arXiv - AI
·
4 min
·
Machine Learning
[2501.00755] An AI-powered Bayesian generative modeling approach for causal inference in observational studies
The paper presents CausalBGM, an AI-driven Bayesian generative modeling approach designed for causal inference in observational studies, ...
arXiv - Machine Learning
·
4 min
·
Machine Learning
[2506.22095] Beyond Simple Graphs: Neural Multi-Objective Routing on Multigraphs
This article presents novel graph neural network methods for multi-objective routing on multigraphs, addressing limitations of existing t...
arXiv - Machine Learning
·
3 min
·
Machine Learning
[2506.05647] Learning to Weight Parameters for Training Data Attribution
This paper introduces a novel method for gradient-based data attribution that learns parameter importance weights from data, enhancing at...
arXiv - Machine Learning
·
3 min
·
Llms
[2408.07238] Beyond Mimicry to Contextual Guidance: Knowledge Distillation for Interactive AI
This article presents a novel approach to knowledge distillation for interactive AI, emphasizing contextual guidance over simple output i...
arXiv - Machine Learning
·
4 min
·
Ai Agents
[2310.01331] ChoiceMates: Supporting Unfamiliar Online Decision-Making with Multi-Agent Conversational Interactions
The paper presents ChoiceMates, a multi-agent system designed to assist users in making unfamiliar online decisions by facilitating inter...
arXiv - AI
·
4 min
·
Machine Learning
[2505.14825] Assimilative Causal Inference
The paper presents Assimilative Causal Inference (ACI), a novel framework that utilizes Bayesian data assimilation to identify dynamic ca...
arXiv - Machine Learning
·
4 min
·
Machine Learning
[2505.14338] Better Neural Network Expressivity: Subdividing the Simplex
This paper investigates the expressivity of ReLU neural networks, demonstrating that fewer hidden layers are needed than previously conje...
arXiv - Machine Learning
·
4 min
·
Llms
[2505.11409] Visual Planning: Let's Think Only with Images
The paper introduces 'Visual Planning', a new paradigm that utilizes images for reasoning in spatial tasks, enhancing planning capabiliti...
arXiv - Machine Learning
·
4 min
·
Machine Learning
[2504.02922] Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning
This article discusses advancements in model diffing using crosscoders to better interpret changes in AI models during chat-tuning, addre...
arXiv - Machine Learning
·
4 min
·
Generative Ai
[2601.06500] The AI Pyramid A Conceptual Framework for Workforce Capability in the Age of AI
The article presents the AI Pyramid, a framework for understanding workforce capabilities in an AI-driven economy, emphasizing the need f...
arXiv - AI
·
4 min
·
Llms
[2512.24008] SPARK: Search Personalization via Agent-Driven Retrieval and Knowledge-sharing
The paper presents SPARK, a framework for personalized search using agent-driven retrieval and knowledge-sharing, enhancing user experien...
arXiv - AI
·
4 min
·
Machine Learning
[2512.15783] AI Epidemiology: achieving explainable AI through expert oversight patterns
The paper presents 'AI Epidemiology', a framework for enhancing explainability in AI systems through expert oversight, using population-l...
arXiv - Machine Learning
·
4 min
·
Ai Safety
[2511.11924] A Neuromorphic Architecture for Scalable Event-Based Control
This paper presents a neuromorphic architecture for scalable event-based control, leveraging the rebound Winner-Take-All motif to integra...
arXiv - AI
·
3 min
·
Nlp
[2511.10164] Two Constraint Compilation Methods for Lifted Planning
This paper presents two innovative constraint compilation methods for lifted planning in AI, addressing scalability issues in existing co...
arXiv - AI
·
3 min
·
Related Topics
Stay updated with AI News
Get the latest news, tools, and insights delivered to your inbox.
Daily or weekly digest • Unsubscribe anytime