Large Language Models

arXiv:2601.09195v2 Announce Type: replace-cross Abstract: Supervised fine-tuning (SFT) is a fundamental post-training strategy to align Large Language Models (L

arXiv:2601.10305v3 Announce Type: replace-cross Abstract: Vision-Language Pre-training (VLP) models have achieved remarkable success by leveraging large-scale i

arXiv:2601.11702v2 Announce Type: replace-cross Abstract: AI compliance is becoming increasingly critical as AI systems grow more powerful and pervasive. Yet th

arXiv:2601.19072v2 Announce Type: replace-cross Abstract: Large Language models (LLMs) have shown strong capabilities in code review automation, such as review

arXiv:2602.02378v2 Announce Type: replace-cross Abstract: As LLMs expand from assistance to decision support, a dangerous pattern emerges: fluent agreement with

arXiv:2602.07058v2 Announce Type: replace-cross Abstract: Machine Unlearning aims to remove the influence of specific data or concepts from trained models while

arXiv:2602.07150v3 Announce Type: replace-cross Abstract: Agentic systems are evaluated on benchmarks where agents interact with environments to solve tasks. Mo

arXiv:2602.07906v4 Announce Type: replace-cross Abstract: Autonomous Machine Learning Engineering (MLE) requires agents to perform sustained, iterative optimiza

arXiv:2602.16485v2 Announce Type: replace-cross Abstract: Existing Multi-Agent Systems (MAS) typically rely on homogeneous model configurations, failing to expl

arXiv:2602.19345v2 Announce Type: replace-cross Abstract: Group Relative Policy Optimization (GRPO) has significantly advanced the training of large language mo

arXiv:2602.23811v3 Announce Type: replace-cross Abstract: We investigate the theoretical aspects of offline reinforcement learning (RL) under general function a

arXiv:2603.09643v3 Announce Type: replace-cross Abstract: Current evaluation frameworks and benchmarks for LLM powered agents focus on text chat driven agents,

arXiv:2603.11066v3 Announce Type: replace-cross Abstract: We develop a structural and quantitative framework for analyzing the Collatz map through modular dynam

arXiv:2603.12510v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models have significant potential to enable general-purpose robotic syste

arXiv:2603.12564v3 Announce Type: replace-cross Abstract: Tool-augmented LLM agents increasingly serve as multi-turn advisors in high-stakes domains, yet their

arXiv:2603.13119v2 Announce Type: replace-cross Abstract: Camera motion is a fundamental geometric signal that shapes visual perception and cinematic style, yet

arXiv:2603.14867v2 Announce Type: replace-cross Abstract: Many strategic decision-making problems, such as environment design for warehouse robots, can be natur

arXiv:2603.15970v3 Announce Type: replace-cross Abstract: Several data warehouse and database providers have recently introduced extensions to SQL called AI Que

arXiv:2603.17872v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have achieved unprecedented fluency but remain susceptible to "hallucinat

arXiv:2603.18385v2 Announce Type: replace-cross Abstract: We present a new solution concept called evolutionarily stable Stackelberg equilibrium (SESS). We stud

arXiv:2603.20957v2 Announce Type: replace-cross Abstract: Frontier LLM companies have repeatedly assured courts and regulators that their models do not store co

arXiv:2603.21576v2 Announce Type: replace-cross Abstract: Long-context LLM inference is bottlenecked not by compute but by the O(n) memory bandwidth cost of sca

arXiv:2603.21853v2 Announce Type: replace-cross Abstract: This paper proposes a novel alternative to existing sim-to-real methods for training control policies

Qwen3.5-9B-Uncensored-HauhauCS-Aggressive is an uncensored variant of the base model created by Hauhau CS. This 9-billion parameter model removes safety filters

In this post, we explore how the multimodal foundation models (FMs) of Amazon Bedrock enable scalable video understanding through three distinct architectural a

The first lady sees AI and robotics playing a prominent role in the future of American education.

In this series of posts, you will learn how streaming architectures help address these challenges using Pipecat voice agents on Amazon Bedrock AgentCore Runtime

With female AI fruit being fart-shamed and even sexually assaulted, there’s a misogynistic undercurrent to the fruit slop microdramas, even as they appear to be

In a controlled experiment, OpenClaw agents proved prone to panic and vulnerable to manipulation. They even disabled their own functionality when gaslit by huma

Spotify is beta-testing a new feature called Artist Profile Protection that lets artists review releases before they go live. Sometimes songs end up on the wron

Google DeepMind researches AI's harmful manipulation risks across areas like finance and health, leading to new safety measures.

Granola, the London-based AI meeting app that records conversations without dropping a bot into the call, has raised $125 million in a Series C round led by Dan

China begins to take over the open source AI race as models like DeepSeek and Qwen see greater adoption among developers.

Learn to build an AI data analyst with Python: analyzes data, detects anomalies, and generates predictions using local LLMs.

Fleet now supports shareable skills, so you equip agents across your team with knowledge for specialized tasks.

Yes, you can use AI cost-effectively. Here's how professionals are doing it.

Less than a week into his tenure as Disney's newly-appointed CEO, Josh D'Amaro is already dealing with two separate crises that have cast a shadow over the comp

Introducing Lyria 3 Pro, which unlocks longer tracks with structural awareness. We’re also bringing Lyria to more Google products and surfaces.

Lyria 3 is now available in paid preview through the Gemini API and for testing in Google AI Studio.

We are bringing Lyria 3 to the tools where professionals work and create every day.

Humanoid robots are attracting capital at a pace the underlying technology cannot yet justify. Between viral dance performances, stratospheric valuations, and f

Investors like Sequoia, Andreessen Horowitz, Kleiner Perkins, and Elad Gil can't get enough of AI legal tech startup Harvey.

context engineering is about designing the information ecosystem that the model has access to when it processes your request. Faros AI identified eight layers o

As the ChatGPT-maker eyes an IPO, it's ditching Sora in favor of a unified AI assistant and enterprise coding tools.

Granola's valuation jumped from $250 million to $1.5 billion with this round, and it has added more support for AI agents after users previously complained.

Meta is using generative AI to provide more product and brand information to consumers when they're shopping in its apps.

Last September, when OpenAI quietly released the Sora 2 app to the public, the discourse around it was not quiet at all. Commentators who had spent months watch

My friend who is a developer once asked an LLM to generate documentation for a payment API.