AI Tools & Frameworks Directory

arXiv:2602.04902v1 Announce Type: new Abstract: The Mechanistic Interpretability (MI) program has mapped the Transformer as a precise computational graph. We extend this graph with a conservation law and time-varying AC dynamics, viewing it as a physical circuit. We introduce Momentum Attention, a ...

arXiv:2602.04903v1 Announce Type: new Abstract: Feature steering has emerged as a promising approach for controlling LLM behavior through direct manipulation of internal representations, offering advantages over prompt engineering. However, its practical effectiveness in real-world applications rem...

arXiv:2602.04904v1 Announce Type: new Abstract: Multimodal fusion faces two robustness challenges: noisy inputs degrade representation quality, and missing modalities cause prediction failures. We propose DCER, a unified framework addressing both challenges through dual-stage compression and ener...

arXiv:2602.04986v1 Announce Type: new Abstract: We endorse and expand upon Susan Schneider's critique of the linear model of AI progress and introduce two novel concepts: "familiar intelligence" and "strange intelligence". AI intelligence is likely to be strange intelligence, defying familiar patte...

arXiv:2602.05014v1 Announce Type: new Abstract: With the rapid progress of tool-using and agentic large language models (LLMs), Retrieval-Augmented Generation (RAG) is evolving from one-shot, passive retrieval into multi-turn, decision-driven evidence acquisition. Despite strong results in open-dom...

arXiv:2602.05059v1 Announce Type: new Abstract: Large Language Models are increasingly used by students to explore advanced material in computer science, including graph theory. As these tools become integrated into undergraduate and graduate coursework, it is important to understand how reliably t...

arXiv:2602.05073v1 Announce Type: new Abstract: Uncertainty quantification (UQ) for large language models (LLMs) is a key building block for safety guardrails of daily LLM applications. Yet, even as LLM agents are increasingly deployed in highly complex tasks, most UQ research still centers on sing...

As trust in the old order erodes, mid-sized countries are building new agreements involving shared digital infrastructure and localized AI.

The startup -- with backing from Accel -- is building a financial layer that handles the authentication and micro-payments required for AI agents.

Anthropic has launched Claude Opus 4.6, its most capable model to date, focused on long-context reasoning, agentic coding, and high-value knowledge work. The model builds on Claude Opus 4.5 and is now available on claude.ai, the Claude API, and major cloud providers under the ID claude-opus-4-6. Mod...

The new model is build to accelerate the capabilities of Codex, the agentic coding tool OpenAI launched earlier this week.

OpenAI has just introduced GPT-5.3-Codex, a new agentic coding model that extends Codex from writing and reviewing code to handling a broad range of work on a computer. The model combines the frontier coding performance of GPT-5.2-Codex with the reasoning and professional knowledge capabilities of G...

OpenAI launched Frontier, a new platform designed for enterprises to build and deploy agents while treating them like human employees.

The newest version of Anthropic's model is designed to broaden its appeal.

Liang Mou | Staff Software Engineer, Logging PlatformYisheng Zhou | Software Engineer II, Logging PlatformElizabeth (Vi) Nguyen | Software Engineer I, Logging PlatformOwen Zhang | Senior Software Engineer, Logging PlatformIntroductionAs Pinterest has grown, the demand for a robust, real-time, and co...

A collage of four images, the first of a woman with curly hair in front of a silver laptop, the second of the same woman and a man with short black hair speaking on a stairwell, the third of a the same man with glasses, and an aerial image of NTID

This research looks at why Predictive Inverse Dynamics Models often outperform standard Behavior Cloning in imitation learning. By using simple predictions of what happens next, PIDMs reduce ambiguity and learn from far fewer demonstrations. The post Rethinking imitation learning with Predictive Inv...

Stop fighting AI. Use a tech stack AI understands and can build a paid SaaS within minutes.

Fundamental has built a new foundation model to solve an old problem: how to draw insights from the huge quantities of structured data produced by enterprises.

Are the human-like cognitive abilities of LLMs real or fake? How does information travel through the neural network? Is there hidden knowledge inside an LLM? The post Mechanistic Interpretability: Peeking Inside an LLM appeared first on Towards Data Science.

Joint collaboration unites powerful robotics with AI-driven orchestration to redefine how labs scale automation HighRes, Inc., a leader in laboratory automation and orchestration software, and Opentrons Labworks, Inc., the leading laboratory robotics company building the physical infrastructure for ...

NTT DATA, a global leader in AI, digital business and technology services, today announced a multi-year Strategic Collaboration Agreement (SCA) with Amazon Web Services (AWS) to help enterprises modernize legacy systems, adopt agentic AI responsibly and scale innovation across industries. Combining ...

Intruder, a leader in exposure management, today released new security research detailing vulnerabilities in Moltbot, formerly known as Clawdbot, an open-source, self-hosted AI assistant. The research, “Clawdbot: When Easy AI Becomes a Security Nightmare,” finds that Moltbot’s emphasis on rapid, sim...

Fortune 300 leader is deploying AI agents on Snowflake to help teams across 1,600+ branches make faster, data-driven decisions Snowflake (NYSE: SNOW), the AI Data Cloud company, today announced that United Rentals, widely recognized as the world’s largest equipment rental provider, has rolled out a ...