Maayan Roth, a software engineer, is revisiting machine learning fundamentals by reading Andriy Burkov's 'The Hundred-Page Languages Models Book' to better understand large language models. In Chapter…
Researchers have released Tpo-Torch, a PyTorch implementation of Target Policy Optimization (TPO), a reinforcement learning from human feedback (RLHF) algorithm that simplifies the standard PPO approa…
HuggingFace suffered an AI-assisted cyberattack last week and was unable to analyze it using frontier models from Anthropic or OpenAI because those models' safety guardrails blocked cybersecurity-rela…
On July 16, 2026, HuggingFace disclosed a security breach in which an autonomous AI attacker infiltrated its internal infrastructure by chaining two remote code execution vulnerabilities in its datase…
HuggingFace disclosed in its July 2026 security incident report that safety guardrails on commercial AI models blocked its own forensic analysis of a breach, forcing the team to switch to the open-wei…
Mlx-serve, an open-source native LLM inference server for Apple Silicon written in Zig, shipped an update that adds compatibility with the Ollama API alongside existing OpenAI and Anthropic API suppor…
Moonshot AI released Kimi K3 on July 16, 2026, a 2.8-trillion-parameter Mixture-of-Experts model that ranks #3 on the Artificial Analysis leaderboard and #1 on Arena.ai for front-end coding, but the m…
University of Illinois Urbana-Champaign researchers have released an open-source AI teaching assistant on HuggingFace that runs 11 separate models in parallel for text/image retrieval, generation, mod…
Moonshot AI released Kimi K3, a 2.8-trillion-parameter mixture-of-experts model with 16 active experts per token, achieving the top score of 1679 on the Artificial Analysis webdev arena ahead of Claud…
Thinking Machines Lab, the startup founded by former OpenAI CTO Mira Murati, launched Inkling, an open-weight 975-billion-parameter large language model that lets users adjust thinking effort. The mod…
Thinking Machines has released Inkling, a 975B-parameter open weights model with 41B active parameters, debuting at 41 on the Artificial Analysis Intelligence Index and becoming the leading open weigh…
A security challenge called 'The Hoard' offers $51,200 to anyone who can find a backdoor hidden in one of seven open-source language models on HuggingFace, with the bounty doubling daily until July 21…
A new data provenance system called ob enables record- and token-level tracing for precise data deletion, reducing over-deletion from 101x to 1.3x in tests on 219,555 Wikipedia pages. The system adds …
Researchers present OriginBlame (ob), a record- and token-level data provenance system that enables precise identification of training records belonging to a specific author, addressing the gap betwee…
Meta and MLC AI have released LlamaFile, a portable binary format that bundles an entire large language model with its inference engine into a single executable file, enabling users to run 100+ open-s…
A developer built a Retrieval-Augmented Generation (RAG) system from scratch using Python, LangChain, Ollama, and Chroma, demonstrating how to combine retrieval and generation for enterprise AI assist…
A developer's analysis quantifies the real operational costs of open-source AI tools in agent-centric architectures, revealing that while models like LLaMA 3 are free to use, deployment can cost over …
A new study reveals that the multiplicative repetition penalty used across major LLM inference engines (HuggingFace, vLLM, llama.cpp) is fundamentally flawed because it branches on the sign of raw log…
A developer known as P-Kumar-18 is building Siagnos, a personal taste engine for fanfiction recommendations that models individual reading preferences rather than relying on popularity metrics. The pr…
Four months after Google Research and DeepMind published TurboQuant, a training-free algorithm suite claiming 6x memory reduction for KV cache, the promised official code has not shipped. Instead, the…