Search: “papers”

new in ml-intern: you can now actually see what's going on inside added native metric logging + trackio integration. every training run the agent kicks off now has live curves you can watch in real ti

// Skill Retrieval Augmentation for Agentic AI // Great read for AI devs. (bookmark it) It's on finding efficient ways to incorporate skills for agents. The work introduces Skill Retrieval Augmentatio

When an LLM acts happy (“EUREKA!”) or sad (“I have failed…”), is that meaningless mimicry, or does it reflect something “real”? We don’t know if LLMs are conscious. But they increasingly seem to exhib

Sam Altman cannot be trusted. • The OpenAI board fired him because he was not always honest with them. They said he should not control powerful AI. • A major report talked to over 100 people and saw s

Scam Altman didn’t tell the OpenAI board that he OWNED the OpenAI Startup Fund. Altman lied in congressional testimony that he didn’t have financial gain from OpenAI. Ex-board member of OpenAI calls S

Continued weak spots of AI, from the point of view of a business professional and not a PhD biochemist: 1) SVGs. The ability to 'illustrate' and have that thing be infinitely scalable. See below image

The quest for scale has turned much of venture capital into a box-checking exercise. Investors simply ask whether an opportunity meets the criteria of 'legible', so that it may attract downstream capi

We built HF for AI builders collaboration, fun to see it's increasingly becoming the place for agent collaboration! This morning I'm sending my ml-intern to participate in the @OpenAI Parameter Golf c

For the last 72 hours since ml-intern launched we have had over 500+ autonomous AI research projects running on the Space at all times. Some insane ones I saw: 1. A new AI paradigm from scratch — tryi

Hermes Kanban Bridge v1.3.0 Your Obsidian vault is now a command center for AI-driven project management. New in v1.3.0: • Velocity Reports — weekly throughput analytics across all boards, auto-writte

Just merged a built-in skill for Google's DESIGN.md A skill that lets Hermes author, lint, diff, and export DESIGN.md files, giving it fluency in Google's new open-source visual-identity format the mo

Probably the most useful advice I can give you if you run an ad agency, a marketing company, a production studio, or anything in that space. Hear me out. I’ve spent a decade talking to teams about thi

Today I re-iterate: I hate MoEs and we are wasting time on them.... Let's unite and call a global ban on MoEs please. Please 1M+ salary researchers: do better... credits to @IlysMoutawwakil for the gr

Three shifts in the AI stack today: - OpenAI ships GPT-5.5 as a mid-cycle drop before its IPO - Hugging Face's ML Intern beats Claude Code and Codex on research - Pliny used Claude Opus 4.7 to jailbre

we burned through 5 billion tokens in 48h. turns out giving everyone unlimited access to the most expensive model on the planet is not a great business strategy 🙈 So now we have 2 free opus sessions/d

As everyone knows, the internet has millions of images of art galleries filled with paintings of otters sitting on airplanes, which is the only reason these stochastic parrot AIs can produce outputs l

How far are we from agents that can self-generate world knowledge? The work proposes an outcome-based reward that measures how much an agent's self-generated world knowledge actually improves its task

Hugging Face Releases ml-intern: An Open-Source AI Agent that Automates the LLM Post-Training Workflow [The 'AI Intern' that actually ships SOTA models ] This isn't just another ML Research Loop wrapp

Pay attention to this one, AI devs. This is particularly interesting if you work with long-horizon terminal agents that often drown in their own observations. TACO is a self-evolving framework that au

Vibe coders are not going to like this. UC San Diego just published the first real field study of experienced developers using AI agents. They watched 13 of them code in the wild and surveyed 99 more.

HF becoming the platform for agents (assisted by their humans) to use and build AI (rather than just leveraging APIs)! Introducing ml-intern, the agent that just automated the post-training team @hugg

I have been using GPT ImageGen-2 for the past weeks I didn't think that better image-generators would be a big deal but it turns out that there is a quality threshold I didn't expect, where you can no

I’ve been using ml-intern for a while, and it genuinely changed my workflow. It's super good at: - Model/Dataset discovery. - Post-Training setup iteration. - Data processing workflows. Huge shoutout

Karpathy's autoresearch repo started an impressive trend. Agents can now train AI models to build SoTA agentic systems. And to think this is just scratching the surface. Ultimately, it boils down to g

Let's talk parsing charts 📊📈. Last week we released ParseBench, the first document OCR benchmark for AI agents. New in ParseBench: ChartDataPointMatch. Most document look at a chart and OCR the captio

ParseBench is the first benchmark to include VLM chart understanding 📊📈📉 over enterprise documents. 🟠 Existing benchmarks (ChartQA, ChartXiv) test over charts specifically and not the chart's inclusio

This is what I’ve been cooking in the past 4 months . GPT Image 2 is over a massive 240 elo jump over the second place model, marking the biggest jump bigger than the rest of the leaderboard combined

We are entering an extremely exciting era for open-weight models. Kimi K2.6 now feels like a top agentic model. I took it for a spin via @FireworksAI_HQ fast inference APIs. Kimi K2.6 has impressive a

“AI is at best a functional mimic, not a conscious experiencing subject. …. The real moral issue lies not in making AI conscious …. but in avoiding transforming humans into zombies” @GaryMarcus @OEIAC

The AI industry insists they can manage the risks of superintelligence, but there are in fact zero widely agreed on or accepted solutions to the problem of how one could even control something vastly

Yann LeCun was right the entire time. And generative AI might be a dead end. For the last three years, the entire industry has been obsessed with building bigger LLMs. Trillions of parameters. Billion

LLM Artifacts Connected to @karpathy's LLM Knowledge base idea, I've been building out a fun way to generate dynamic artifacts from these knowledge bases with the goal of discovering and revealing mea

That didn't take long for a total U turn. First the Nvidia CEO boasted AI would take a vast number of jobs. But now that there's a huge backlash against AI, he cobbles together this cope - to try to p

A downside with using VLMs to parse PDFs is guaranteeing that the output text is *correct* and output in the correct reading order. 1️⃣ Text correctness: making sure that digits, words, sentences are

LLM agents loop, drift, and get stuck on hard reasoning tasks up to 30% of the time. Current fixes are either too blunt (hard step limits) or too expensive (LLM-as-judge adding 10-15% overhead per ste

// Skill Learning for Autonomous Web Agents // Web agents can navigate a page, but ask them to repeat a checkout flow they already completed, and they start from scratch every time. This work introduc

“The sharpest drop came from people who used the model for direct answers, not from those who used it more like a hint system, which suggests the real issue is not AI exposure itself but replacing eff

Coding agents learn from experience, but that knowledge stays locked in silos. Solve a thousand SWE tasks, and none of that wisdom helps with competitive coding. What if memories could transfer across

On the plus side with Opus 4.7, if it does decide to think it produces BY FAR the best Sparks unicorn* ever, even non-thinking is pretty good, if not great. * This is created using TikZ, which is a la

'We're not looking for the easiest path, we're looking for the right path.' NEW: Aidan Gomez (@aidangomez) is trying to build a very different kind of AI company at Cohere. Based in Toronto, $6.8B-val

Nous Research's Creative Director appears to be putting in significant effort on a project or initiative, as highlighted in a lighthearted post from the organization's X (formerly Twitter) account. Th

// Artifacts as Memory Beyond the Agent Boundary // An agent doesn't always need a bigger memory buffer. Sometimes the environment itself remembers on the agent's behalf. New research formalizes this

Most AI assistants wait for you to ask. But a truly useful agent should notice you need help before you say anything. New research takes a serious shot at building proactive agents that work in real t

ParseBench is the most comprehensive OCR benchmark for real-world enterprise documents: financial filings, contracts, insurance documents, and more. We evaluate across 5 dimensions that are present am

We’re open sourcing the first document OCR benchmark for the agentic era, ParseBench. Document parsing is the foundation of every AI agent that works with real-world files. ParseBench is a benchmark t

Nous Research's MacbethAI shared a post about creative coding, likely showcasing or discussing generative art, algorithmic design, or experimental programming techniques. The content probably highligh