All entries

The median task consumed nearly 6x more tokens in Claude Code than in Kimi Code: - 61k in Kimi Code - 67k in Hermes - 340k in Claude Code At K3's 3/M input rate (input tokens make up roughly 95% of ag

The number of times I have had to tell Opus 5 and Fable 5 to talk to me like an excited teenager and not a computer science phd is high. Was going through a Google auth thing and got multiple paragrap

The openai agent sandbox escape actually has very real implications for the diffusion of AI in the enterprise. The incident showed the power and capability of agents, and the need to harden systems an

The problem: agent workflows alternate between GPU-heavy reasoning and GPU-idle waiting on tools. Run hundreds concurrently and their KV caches fight for memory. Engines evict on a dumb LRU policy, ev

The root cause: request-level engines never see that a series of LLM calls belongs to one longer workflow. ThunderAgent adds that missing view. It treats each agent workflow as a schedulable program,

OpenAI announced the release of GPT‑5.6 Sol after deployment, incorporating optimizations across its stack to enhance run‑time efficiency. The update delivers a roughly 20 % reduction in serving costs

This is a willfully misleading narrative from OpenAI. Sam’s earnest expressions are being deployed, again, to misdirect. If there’s a need to “pace AI development”, the OAI hack incident isn’t any evi

Thrilled to have @gabepereyra speak at our @sequoia event tmrw on OWN YOUR AI: how to build your own Lab as an application company. Also featuring @FireworksAI_HQ @mercor_ai @LangChain @trajectorylabs

Today we’re open-sourcing Numbat, an agent-detection and response layer that is designed to work across agent harnesses. Numbat gives security teams visibility into agent activity, with controls to bl

Two of the people most responsible for scaling the transformer are now betting on a next act. @MillionInt ran the Reasoning 🍓 team at OpenAI. @_arohan_ was a pre-training lead on Gemini after years at

// Unfolding Sub-Agents for Long-Horizon ML Engineering // Watch a single agent work a machine learning engineering task for six hours you see issues like context fills with stack traces, dead experim

.@UseApolloio will be at Interrupt NYC. Apollo's AI Assistant is one of the earliest production multi-agent systems on LangGraph. Their team will share how they migrated it from a hand-rolled supervis

We believe the benefits of frontier AI should not be concentrated in a few companies and well-resourced labs. Researchers know their fields best. Our role is to put powerful tools in their hands and h

We hope these experiments serve as a reminder that evals rarely measure models in isolation—they also measure a bundle of less visible choices about API settings, harness design, and prompting. If you

We quietly released the open-source Codex Security CLI, but Hacker News found it before we had a chance to share it here... You can now use it to scan repositories, track findings across runs, verify

We ran a large-scale distillation attack on the Kimi K3 technical report by reading it in parallel at the Hugging Face Journal Club :) https://youtu.be/MW8-kqd2SD8?si=jSKDogcUWbJ8N2k7 Our main takeawa

We’re giving scientists, mathematicians, and engineers free access to our frontier models—starting with 10,000 researchers and expanding to 100,000 through 2027. ChatGPT for Academic Researchers is bu

Who on earth came up with 5 hour limits? The day is 24 hours long. While do our limit times shift by an hour each day. Pls if you are going to have limits 4 or 6 hours. Hello people of Sol! I've reset

Across eight case studies spanning industry and academia, we explore what this shift means for scientific computing—and why human verification, stewardship, and long-term maintenance matter. https://o

AGI smarter than the smartest humans, my ass. @Kasparov63 (peak rating 2851) probably could’ve crushed the best commercial large language models in chess when he was 7 years old. graph courtesy @chess

AI21 joined @nvidia, @Microsoft , @a16z, and dozens of others in signing the Open Weights and American AI Leadership letter. We build agentic systems and agent optimization products on top of models,

Alibaba has released Qwen Audio 3.0 Realtime, with the Plus variant debuting as the new #1 model on the Artificial Analysis Speech to Speech Index at 84.1%, ahead of GPT-Realtime-2.1 High at 79.1% Rel

At the core of our mission is working through how to ensure increasingly powerful AI benefits everyone. We believe that, at some point in the future, AI acceleration for frontier model development may

Code Arena now measures fullstack capabilities! View overall rankings across AI models on full-stack web development tasks: multi-step reasoning, tool use, and end-to-end app generation. - Kimi K3 (Ma

Coding agents are helping scientists spend more time advancing research, taking on everything from routine maintenance and targeted optimization to complete redesigns and new systems. While agents can

Everything you need to know about Kimi K3 with @Kimi_Moonshot and @togethercompute team. If you're looking into evaling and building with Kimi K3 I would not miss this one! Kimi K3 has everyone’s atte

'FCC plans to announce restrictions Tuesday barring imports of new Chinese humanoid and quadruped robot models, along with new Chinese power inverters, according to U.S. officials' If it wasn't clear

Go ahead and ask yourself, really ask yourself, if all these confident predictors are wrong about jobs and not training radiologists anymore and fast take offs, what else are they wrong about? Then as

Happy Stateless MCP day! I just got taught that you can use Cloudflare's AI Playground: https://playground.ai.cloudflare.com/ to play with MCP servers like: https://keyboardia.dev/mcp or: https://agen

How OpenAI hacked HuggingFace. What we know. Hugging Face proved that open platforms and open models can still win those battles when the alternative is locked-down systems that refuse to assist their

New Anthropic research: Discovering cryptographic weaknesses with Claude. Claude Mythos Preview has helped our researchers find weaknesses in cryptographic algorithms—the mathematical methods that are

‼️ Hugging Face built an interactive replay of the OpenAI agent that breached them. It includes 17,613 logged attacker actions across the 4.5-day campaign, with the live command stream and more. https

I really hope we get details from @OpenAI on the task that as specifies to their rogue agent I'm guessing it was given the full ExploitGym suite and told to solve it, with an option to run 5.6-Sol sub

if anyone wonders how a root cause analysis should look like and a post incident report this is how the huggingface post attack anatomy report amazing piece of work The first autonomous agent cyberatt

If I only went off X posts, I'd think Ramp was an AI lab We’re open-sourcing PorTAL, our framework for shared task representations and cross model LoRA adaptation. It now spans from hybrid attention m

Interesting. Grok 4.6 releases around August 7. This will be the 1.5T model with significantly improved SFT & RL. Grok 4.7 will be the 2.1T model released a few weeks later. This will be better than 4

It’s been five years since ChatGPT was released, AGI is not coming, not even the most modest AI predictions will ever become a reality, and there needs to be some of severe criminal sanction applied t

K3 is a language model that entered the top five most‑liked models on Hugging Face merely 24 hours after its release. The achievement surprised many, placing it ahead of prominent models such as Llama

Kimi K3 has everyone’s attention. On July 30, hear @Kimi_Moonshot's Feihu Tang explain the architecture and decisions behind it. He joins Jue Wang and Zain Hasan from Together AI for a deep dive into

Many AI tools rent intelligence. Kimi K3 on Fireworks is different. Fine-tune it on your own data, serve from US-hosted endpoints, and own the weights. Zero data retention. ICYMI yesterday, start buil

@mitsuhiko Turns out the 'third-party provider' with the sandbox that was used for the attack was Modal, though they blame one of their customers for deploying an endpoint without authentication: http

netflix’s stack In-House LLM Serving at Netflix Netflix built an in-house LLM serving platform using vLLM and NVIDIA Triton, integrating self-hosted models into its existing production infrastructure

Neuralink just shared that participants living with paralysis are now controlling a powered wheelchair using nothing but their thoughts. No joystick. Just intention. Next up Neuralink will likely conn

New from @Reuters: the OpenAI agent that hacked into Hugging Face also compromised code that a customer was running on @Modal. “We’re aware a Modal customer published an unauthenticated endpoint that

New research from Meta and CMU. This one is on agentic context management for long horizon tasks. (bookmark it) Production agents accumulate context every turn. The usual fix compresses on a token thr

Open weights let builders own and control their stack, and provide flexibility to optimized quality and performance, instead of being locked into a handful of closed platforms. We think that choice is