Search: “inference”

Great paper if you are tracking progress in recursive self-improvement (RSI). (bookmark it) There is so much hype around RSI, so I think it's worth understanding why current models are not able to do

This is one of the most effective ways to improve your agentic workflows. If you are building computer-use agents, this one is worth your time. Task Model Induction takes a raw recording of someone wo

What a fascinating paper on AI agents. A lot of the issues we see with AI agents today revolve around wrong assumptions the LLMs make. This leads to problems like hallucination, cost inefficiencies, u

Banger paper from Google DeepMind. It's on the very hot topic of model routing. Routers assume the value estimate is free. Working out which specialist handles a query best costs money too, and it is

Finally a good paper testing whether agents can really post-train other agents. (bookmark it) They analyzed a large corpus of publicly released post-training trajectories. Across tasks, the agent lock

Great paper if you are building production-grade agents. It's a good way to understand what is actually incurring costs in your agents, which involve more complex components and decisions. This works

It is genuinely painful to me that almost nobody — not even doomers who should care deeply — seems to be paying attention to this critical point that might have truly profound consequences for humanit

Regarding Sentence Transformers v6.0 update: A dense model compresses a whole text into one vector, then compares two vectors. A multi-vector model keeps one vector per token, scores every query token

Trending AND fastest-growing on @tryramp for the 2nd straight month. Four months running on the list. Can confirm @arakharazian's macro take is right: spend is shifting to open-weight, cost-saving inf

The amazingly fast (only 30M!) and extremely accurate @answerdotai ColBERT model is now supported in Sentence Transformers to make it easy to create and query embedding indexes locally directly in Pyt

We’re sharing the concrete changes we’re making to strengthen monitoring, security, and alignment as capabilities advance. We’ve introduced stronger workload and network isolation, continuous security

arXiv:2608.14665v1 Announce Type: new Abstract: The temperature that maximizes pass@k is often low for a small sampling budget and higher for a large budget. This pattern has been reported from Codex

A short illustration of how the Claude's watermarking is supposed to work (based on my read of their released materials). In general, when we are generating tokens, there can be multiple high-scoring

Qwen3.8-27B is now part of your everyday life — from smartphones to vehicles. ⚡️🚀Appreciate your work! @MediaTek Congratulations to the @Alibaba_Qwen on the launch of Qwen 3.8! MediaTek continues our

Thanks. The fact that it works so well for ARC is also a clue to the extent of generalization in these models. It was not at all obvious that it would work originally because the model is never traine

Four small architecture decisions can cost up to 47% of a model's long-context performance. New research from Ai2, Carnegie Mellon, and the University of Washington isolates them. Normalization, GQA,

Qwen3.8-Max is live on Together AI on day zero. Proud to have supported early testing alongside @Alibaba_Qwen, @vllm_project, and Inferact to get day-zero support live. We’re continuing to optimize pe

With the smartest person I know, @HarshSensei, we're building evsys-sdk for the community, this open-source repository allows anyone to build their own continual learning system with first-class suppo

DeepSeek V4 Flash 0731 is now available to fine-tune on Together AI. Specialize it for coding, tool use, and your own domain with SFT or DPO, then deploy the fine-tuned model on Together AI for produc

ELON MUSK ON BUILDING A “STAR MIND” POWERED BY THE SUN “We’ve already built the most powerful AI training clusters in the world. And then what we expect to do by the end of next year is about 10 times

Had so much fun giving this talk at @sequoia about @harvey’s moneyball approach to building a research lab. The biggest mistake I made in the early days of Harvey was trying to play the Yankees baseba

How does OpenAI move so quickly? I'm the first to report on friction@openai.com—the AI age's version of a Jeff Bezos question mark email. Here's how it works. https://fortune.com/2026/08/11/openai-emp

Nemotron 3.5 Lightning is available in LM Studio! The model is 30B MoE (3B active), can run very fast, and is trained for high volume agentic use cases. Model page: https://lmstudio.ai/models/nvidia/n

NVIDIA’s Nemotron 3.5 Lightning—a fast open AI model designed for always‑on agents that perform high‑volume, specialized work—has been launched on the Together AI platform as of 11 August 2026. NVIDIA

A lot of the questions we get from developers are about the concepts behind the API: TTFT, context windows, sampling, fine-tuning, quantization, deployment tradeoffs. We added Learn to the Together do

This might be a good time to mention my recent keynotes titled: 'Agentic AI is Neurosymbolic AI' I would have assumed it was fairly obvious, but in case it's not: a million-line codebase (also known a

An accurate characterization of the arc of AI is that it is shaped by two trends: 1. Moving more and more logic to a neural model for tasks where training data can be densely sampled (e.g. the shift f

anyone confused about neurosymbolic AI—and its recent enormous victory—should read this. complete and total vindication for what I have been saying here all along. (see also my essays on Claude Code a

Today, we’re launching Alpamayo 2 Super, our frontier open reasoning model for autonomous vehicles. Beyond seeing, Alpamayo understands and reasons through the complex world - thinks before it acts. I

“poorly evidenced sensationalism makes for bad policymaking, politics, philanthropy and much else.” beautiful coda for a weekend in which people went completely nuts without even asking for a control

New research from Google DeepMind. (bookmark it) SkillSmith treats model weights as an additional modality the LLM reads natively. The augmented model ingests existing prefix weights alongside rich te

To address the limits of deep learning and avoid stalling, the field of AI started by applying patch (1), which started being demoed 9 months later in December 2024 and has now become completely ubiqu

Kimi K3 has set a new bar for OSS model intelligence! 2.8T params, 1M context, OpenAI-compatible API. Complete guide to running Kimi K3 on Together AI 👇 👏👏 @Kimi_Moonshot 👏👏 https://www.together.ai/bl

The “AGI-is-near” community keeps committing the same logical fallacy over and over; I have seen it at least half a dozen times today alone. Every time there’s an advance, I see the same error. Here’s

OPENAI IS GOING TO TAKE THIS ENTIRE MARKET DOWN WITH IT And you don't have to own a single share to get hurt. What I'm about to explain should worry anybody who thinks they're diversified: OpenAI is a

verbalizing one of those aha moments i had that seems retroactively pretty obvious: if you prioritize pretrain data quality enough that commoncrawl isn't good enough for you, you have to build a Whole

Our partners at @depthfirstlabs just released dfs-large1, a specialized model built for finding and validating real vulnerabilities in large enterprise codebases. We helped them to scale the training

Starting in 15 minutes 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 A

i'll say it plainly, hermes agent desktop is the best agentic app i've used, and i'm a little mad i didn't find it sooner. it auto see the models i'm serving, laguna s 2.1 sitting on my dgx spark and

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

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

Our users have been really excited to try Kimi K3 and we're excited to bring it to Augment's product family. It is the most capable open-source model that we have tested to date. Try it today in Cosmo

Happy to have @FireworksAI_HQ as our day0 launch partner and bring Kimi K3 to more developers. With Fireworks, you can now deploy and fine-tune the 2.8T Kimi K3 model with just a few clicks! Kimi K3 i

**Summary:** Kimi K3 is a publicly available 3‑trillion‑parameter language model now live on Fireworks, offering 1 million-token context windows with native vision and reasoning that rivals top closed

We've open-sourced MoonEP, our high-performance communication library for distributed MoE workloads. Built to make expert-parallel communication more efficient at scale, MoonEP helps reduce communicat

everyone is talking about open-weight models. at @huggingface we're invested in open source AI. like, a lot. and more every year. i pulled the public github numbers for our orgs. good excuse to introd