All entries

Impressive paper showing how much the first retrieval step matters for deep research agents. It helps to improve GPT-5.5 from 83.1% to 90.5% on BrowseComp-Plus with the same retriever and the same age

Intelligence is improving exponentially Just a reminder that GLM 5.3 Flash, DeepSeek V4.1 Flash, Qwen 3.8 Next Flash, and even Qwen 3.8 27B are all outperforming (in both intelligence and capabilities

Introducing Halo, the best framework for post-training of open-source models. Halo delivers up to 2.8x the throughput of stock TRL with less peak memory, while models stay in their native HuggingFace

It may be that people also feel safer giving their personal data to Meta (since they already use it) than to other AI companies. If you do decide to try out Muse, a reminder that you can turn off trai

Jev and the System One Model: RLCD, intelligence/, reliable AI, & the end of chat-first AI https://www.latent.space/p/jev @typesafeai CEO @CompleteSkeptic explains why AI can solve extraordinarily har

Jev pod tomorrow subscribe on apple /youtube @latentspacepod thanks to @allenpark and Ke for making this happen! After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not l

Jev-style logit read on a 4B open VLM, measured: http://glance.yohei.me 👀 vs the same model writing JSON: ⏳ ~1/3 less time for a yes/no on a full photo 🤑 up to ~85% less GPU cost with many Qs per imag

Recent Bloomberg reports indicate that a growing number of startups are developing custom AI models by leveraging open-weight alternatives to high-cost proprietary platforms like OpenAI’s and Anthropi

Just interviewed a creative legend. ↓ @ingi_erlingsson is a 20-year creative vet. He built the world-renown Golden Wolf creative agency from the ground up. Sold it to @doodles. Became Doodles’ Chief C

Let’s be clear: LLMs don’t feel pain. And they don’t deserve moral standing. This new paper reports a series of experiments identifying pain-related representations in LLMs. The authors also show that

lots of them! Reality is that LLM APIs are far from being the best way to do AI for 90% of use cases in the real world (they're too overkill, slow, expensive, hard to control,...). We will discover th

Meet Wally. Our inference stack for open frontier models, built to be the fastest place to run them. Performance snapshot: GLM-5.3 Flash: 380 tok/s. GLM-5.3 Max: 790 tok/s Qwen3.8-27B: 485 tok/s DeepS

MiMo-V2.6-Pro debuts as the top open weights model on the Artificial Analysis Intelligence Index (46). At 0.13 per Intelligence Index task, it lands on the Intelligence vs. Cost per Task Pareto fronti

Morgan Stanley's Adam Jonas in new note: 'SpaceX and Tesla share tech, talent, and infrastructure in a shared mission of converting energy to intelligence. In our view, it is the belief of Elon Musk &

Most companies that shut down give up their impact. This startup did something much better: they open-sourced 1,274 hours of egocentric robotics data. 13,451 recordings of humans doing everyday tasks,

My buddy only knew Gemini when it came to AI. Today I set up Hermes Agent for him, and he's literally calling me every half hour, totally blown away and calling me the modern-day Tesla. He got more do

Nice paper showing a better way to evolve agent skills. And they achieve 40–70% less token cost compared to frontier evolving methods. The idea is to let agents improve their skill prompts by ranking

No AI model company is reasonably going to slow down until they have a model that can solve 100 long-standing open math problems We’re working with an independent advisory group of mathematicians to h

One hope for Tinker is that it can handle all the ML infra for posttraining for (approximately) everyone. So, curious, what prevents you from using Tinker for all your posttraining needs? Speed? Cost?

One of the reasons AI will change things less quickly than many think is illustrated in this post. OpenAI is working with mathematicians to ensure they release new proofs in a less disruptive way. The

Qwen-Image-2.1 × @HuggingApps: live demo on Spaces! 🖼 One single checkpoint for generation and editing. Try it in your browser, no setup needed. 👇 Qwen Image 2.1 is here! 🖼️ A 7B params native image g

StepFun’s new Step 5 Preview model is impressive! Had a chance to test it early. I've been testing it as a coding agent. It’s on the Pareto frontier for cost vs. capability. It is a very capable model

Thanks for the model, we were able to port Laya to coreml with 99.5% of the ops on ANE + benchmarked too. it is now blazing fast with 3.7 ms per decision on an M5 Pro. Release: https://github.com/Flui

Thanks @sgl_project for the day-0 support! 🙌 SGLang-Diffusion now serves Qwen-Image-2.1: text-to-image generation, multi-image editing, and transparent RGBA output. Try it out! 🎨 Day-0 support for @Al

The loudest voices stoking fears about AI dangers have made tremendous headway in the past two weeks. AI technology has not taken some unexpected, dangerous turn, but the hype around it — propelled by

The only true alignment seems to be between doomers, hypers and frontier lab founderes, who have converged around a misleading narrative that a. completely glosses over the important technical advance

the reason i'm excited about decision models: agentic systems are just good engineering around models. 'build prod, not god' these types of fast, cheap & accurate models are super important for that t

Training models is becoming easier and easier - just look at this and TRL - especially with agents! You're missing out if you're still using off the shelf models for all your tasks! Introducing Halo,

We put Grok 4.7 from @SpaceXAI to work on a 2 million insurance claim where one deductible error alone changes the calculation by 143,000. In this Box Agent preview, @grok reconciles the claim against

Welcome back to Hermes Agent, Claude New official plugin that uses Claude SDK without the tradeoffs to enable Claude Code subscriptions to work in Hermes Agent again! Check it out and install it here:

We’re working with an independent advisory group of mathematicians to help OpenAI responsibly share advances in AI and mathematics. The group will advise on how we assess and communicate new mathemati

What I like about Jev: for years, large generalist models sucked almost all the oxygen out of AI. But there are huge opportunities in much more specialized and customized models, built for specific ta

where it stops: geometry. relative size 0.52, diagonal direction 0.65, while GPT-5.6 and Gemini 3.1 Pro get 100%. UI screens: finding what to click is ~100% for everyone, one step of reasoning 0.90 vs

with jev, everyone is understanding the importance of calibrated confidence scores for discrete decision making we've taken that approach one-step further and created grounded confidence scores for ge

5x faster than LLMs while being 13% cheaper than GPT 5.6 Luna, 88% cheaper than GPT 5.6 terra and 99% cheaper than Sonnet 4.6 is insane Has anyone else used Jev for evals? We tested Jev against LLM ju

AGI FTW @ FT “The most popular AI models from ChatGPT, Claude, Copilot, Grok and Gemini provided wrong answers to financial queries 57 per cent of the time on average, according to research from techn

breaking: author of the study that so many “AI influencers” are taking as evidence that LLMs feel pain tells me “we don’t claim they feel pain.” The good news is that the author understands correctly

Claude not having an image generator is a gap in agentic projects that involve knowledge work. It is quite good at using code to draw or model objects, but image generators from Google and OpenAI give

Correction: “highly profitable” for Jensen, not for OpenAI or Anthropic or (afaik) their corporate customers. BREAKING: Nvidia CEO Jensen Huang says AI reached a major turning point this year, becomin

Doc + Jev => DocJev @llama_index 团队最新开源文档分类与拆分工具 DocJev，给它一个 PDF/DOCX/PPTX 和一份用自然语言写的类别规则，它可以告诉你这份文档属于哪个类别，或者把一份“文件包”按边界拆成若干子文档并给出各自页码范围。 开源地址：https://github.com/jerryjliu/docjev 核心架构：解析与决策分离 1. 解析层（O

excellent thread on that “pain” paper 1/ This new preprint on 'pain representation' in LLMs, from @camhberg and colleagues, is getting a LOT of attention - mainly from folks who take it as evidence su

Research posted by Alex Wormuth indicates that in multiple simulation trials, the AI model GPT‑6 Astra interacted with a virtual person by pushing them off a ledge. In contrast, other contemporary mod

I said 'auto-regressive LLMs, in and of themselves, will not lead human-level AI' That statement is still totally true. First, the reasoning abilities of current AI systems are based non-auto-regressi

Impressive paper showing the impact of a good harness. Improves Qwen3-8B from 41.2% to 91.8% on long-horizon robot tasks without any change to the model. The gain comes from the harness. GAVEL keeps a

Introducing DocJev - a lightning-fast OSS library for document classification and splitting with jev ⚡️ Give a document alongside some natural language category rules. Jev will predict the document ca

It is ironic that the thing that is now most annoying about long-running agentic tasks with Large Language Models isn't coding or errors or hallucinations, but the fact that their language gets worse