Grok is closing the loop on real-world use cases
Grok is closing the loop on real-world use cases @OpenAI released 3 new models yesterday and we immediately tested it on our internal benchmark. All three outperforming gpt-5.5 but @SpaceXAI still the
Knowledge catalogue
Grok is closing the loop on real-world use cases @OpenAI released 3 new models yesterday and we immediately tested it on our internal benchmark. All three outperforming gpt-5.5 but @SpaceXAI still the
the sun is out today To celebrate the launch of GPT-5.6 Sol, we will reset the rate limits again (twice) across ChatGPT Work and Codex over the next 24 hours. We want you to have the time to truly try
This time it is novel math proofs with a public model (most of the other big math breakthroughs have been with experimental LLMs). Yesterday, we made GPT-5.6 Sol Ultra generally available. Today, we'r
Is Meta AI back? Haven't seen Mark post in three years here. Plus, the model is available via API. Not to mention the courage to announce it the same week as the long-awaited GPT-5.6. Great timing if
SpaceXAI's Grok 4.5 takes the #1 spot on AutomationBench-AA with a score of 51%, ahead of Claude Fable 5 (49%) and Claude Opus 4.8 (48%) at roughly a quarter of their cost per task - the first model t
Highly-recommended read. Aligns with what I see in my own harness: > Pi harness got the same success rate as harnesses from the LLM vendors with Opus and GPT, but at 2x less cost > GLM 5.2 was a major
I often use Pinokio to download and work with local open source AI tools/models, and one neat thing is that it exposes them seamlessly to Code and Codex. GPT-5.6 actually found my local video models a
arXiv:2607.05690v1 Announce Type: new Abstract: Language agents run a loop - observe, reason, act - but the memory they reason over sits outside it: a store queried at most once per turn. We study the
nope! OpenAI would not have the nerve. Hey @GaryMarcus, quick question. Did you get months of early access to OpenAI's GPT-5.6 Sol like every AI influencer on X apparently did? Asking because their re
arXiv:2607.04436v1 Announce Type: cross Abstract: Natural language requirements (NLRs) are essential for bridging communication gaps among diverse stakeholders in software development. However, the in
arXiv:2607.03929v1 Announce Type: cross Abstract: Multi-Meta-RAG improves retrieval for multi-hop question answering by filtering a vector store on metadata (the news source) that it extracts from eac
arXiv:2511.07403v2 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) have achieved remarkable progress in vision-language tasks, but continue to struggle with spatial rea
The release of Fable 5 just points to the importance of agent orchestration. You really don't need Fable 5 for most tasks. You can plan with Opus 4.8/Fable 5, execute with GPT-5.5, and design with GLM
so much for recursive self improvement, to the degree that it requires scientific taste the other thing im noticing while working on my research projects is how limited these models are GPT-5.5-xhigh
I wrote about the sqlite-utils 4.0rc1 release a couple of weeks ago. Since we only have Claude Fable on our Max subscriptions for a few more days, I decided to see if it could help me get to a 4.0 sta
arXiv:2512.10485v2 Announce Type: replace-cross Abstract: Vulnerability detection methods based on deep learning (DL) have shown strong performance on benchmark datasets, yet their real-world effectiv
The June edition of my sponsors-only monthly newsletter is out. If you are a sponsor (or if you start a sponsorship now) you can access it here. This month: Claude Fable 5, GPT-5.6, and US export rest
arXiv:2607.00304v1 Announce Type: cross Abstract: The bias-reliability tradeoff conjectures that LLM evaluation systems are constrained in (gamma, H, CV) space, where evaluator coupling (gamma), strat
Yes! Pre-classifying routers are going to result in a lot of bad work because routing is hard and tend to underestimate the value of intelligence on many problems. OpenAI learned this with GPT-5, now
arXiv:2606.29213v1 Announce Type: new Abstract: OCR systems, ranging from classical engines to specialised OCR vision-language models (OCR-VLMs) and frontier multimodal LLMs, report strong results on
GLM-5.2 just tied for #2 on SportEval’s World Cup Exact Score Hit Rate leaderboard. ⚽ 13 exact-score hits out of 76 matches 📊 17.1% accuracy 🥈 Second only to GPT-5.5 That puts @Zai_org ahead of Claude
arXiv:2606.28666v1 Announce Type: cross Abstract: Agent-based AI has enabled the automation of tasks by exposing application tools and resources to large language models (LLMs). However, to improve sc
Row-Bot runs a LangGraph agent at its core. And supports 2 voice pipelines: - Local STT using whisper + Local TTS using kokoro - A real-time voice pipeline using GPT realtime 2. Both support full agen
arXiv:2606.27909v1 Announce Type: cross Abstract: Theory-of-mind evaluations of large language models typically use dyadic social-deduction games, where every observable cue points to a single hidden
Gary’s analysis is pretty much spot on as ever. This is such a worrying development People are not getting how significant this news – The White House delaying GPT 5.6 – is. Here’s the context and a s
arXiv:2606.26454v1 Announce Type: new Abstract: Sphere neural networks have achieved symbolic level syllogistic reasoning without training data, raising the question of where the limit of the scaling
Taps sign We believe in broad access and plan to make GPT-5.6 Sol, Terra, and Luna generally available in the coming weeks. For now, at the request of the U.S. government, we’re starting with a limite
In a joint Fireworks and @Faros_AI evaluation of 211 real engineering tasks, Claude Code + GLM-5.2 beat both Claude Code + Opus 4.8 and Codex + GPT-5.5: - Judge score: 0.568 vs. 0.521 and 0.466 - Time
arXiv:2606.26083v1 Announce Type: new Abstract: Speech conveys information through both words and vocal delivery. We evaluate four leading production realtime voice systems-OpenAI's GPT Realtime 2, Go
I’m headed to San Francisco for AI Engineer World's Fair next week. Two things. I'm giving a talk Tuesday, June 30: 'The unreasonable effectiveness of BM25 for agentic search.' GPT-5 is shockingly goo
arXiv:2606.13189v2 Announce Type: replace Abstract: Prompt-based LLMs are increasingly used for stance detection, but harder examples are not always repaired by clearer instructions, reasoning prompts
arXiv:2606.22269v1 Announce Type: cross Abstract: We investigate the translation quality of current large language models (LLMs) for English-to-Hausa and English-to-Fongbe - two typologically distinct
rumors of new models being delayed 🚨 SCOOP(s): - GPT-5.6 has been delayed and will no longer release this week. New target is ~mid-July. - DeepMind are not satisfied with the current state of 3.5 Pro
arXiv:2606.21861v1 Announce Type: new Abstract: Automated classroom engagement recognition holds substantial promise for scalable learning analytics, yet the suitability of modern Vision-Language Mode
We’re also launching the OpenAI Daybreak Cyber Partner Program with leading security software and services providers. Participating partners can use GPT‑5.5 with Trusted Access for Cyber in the securi
arXiv:2606.11642v1 Announce Type: cross Abstract: How far can we reduce the number of physical keys if we endow an ambiguous keyboard with modern language models? Fewer keys increase hardware design f
but but i thought it was magic? What we learned testing Claude Fable/Mythos 5 on Vending-Bench: > Performance: Makes less money than Opus 4.7 and GPT-5.5 > Alignment: A step back. (Opus 4.8 was better
arXiv:2606.08365v1 Announce Type: cross Abstract: Sparse autoencoder (SAE) features are increasingly used to steer language models, but feature steering is rarely clean: the same intervention can beha
blast from the past 3.5 years ago; some things have changed (esp. coding and math, via neurosymbolic techniques) but many haven’t: Bottom line: From the outset Large Language Models like GPT-3 have gr
Introducing Harness-1, a 20B search agent trained with a state-externalizing harness. > frontier-level long-horizon search, rivaling Opus-4.6 and outperforming GPT-5.4 > Context-1-level cost and laten
locked n loaded🔒 Made in @comfyUI, @AdobeAE + @Photoshop, @suno Vid models: @Alibaba_Wan 2.2, @ltx_model 2.3 @thesystms FLW Img models: @openai GPT 2.0, @googleai Nano Banana 2, @krea 2, @grok GFX: @t
arXiv:2511.05722v3 Announce Type: replace-cross Abstract: Large language models (LLMs) such as GPT-5 and Gemini 3 have pushed the frontier of automated reasoning and code generation. Yet current bench
See also: https://x.com/emollick/status/1799046218917822682?s=20 Alignment, of a sort: this paper conducts what they call a “moral Turing Test,” asking people to compare GPT-4o to humans on ethical qu
arXiv:2606.03626v1 Announce Type: cross Abstract: Vision-language models (VLMs) have been explored for visual programming, where they generate code to solve visual tasks. However, most prior work focu
Release: datasette-agent-micropython 0.1a0 I want Datasette Agent to be able to generate and execute Python code safely. This alpha is looking promising so far. GPT-5.5 has so far failed to break out
arXiv:2601.09696v2 Announce Type: replace Abstract: LLMs are increasingly being integrated into clinical workflows, yet they often lack clinical empathy, an essential aspect of effective doctor-patien
arXiv:2606.00660v1 Announce Type: new Abstract: Agentic search requires language model agents to explore many sources and answer complex information-seeking questions. Scaling test-time compute is a p
I wish the logos and textbox-at-the-bottom interfaces for Discord and Codex did not look so alike at a glance. I have confused the two a couple of times, leading to a confused GPT-5.5 and a confused g
arXiv:2605.31294v1 Announce Type: new Abstract: Recent advances in Audio-LLMs like GPT-4o have ushered in an era of conversational interaction with language models. Conversational avatars however, sti
May ships trace-based evaluation for any agent on any cloud, Grok 4.3 and DeepSeek V4 in the model catalog, GPT-5 Reinforcement Fine-Tuning at gated GA, three Microsoft Research on-device agent models
Aleph 2.0 Introducing Runway Aleph 2.0. Edit videos using AI, keeping camera and motion all consistent. Select specific frames in your video to reprompt with GPT Images 2.0 or Nano Banana 2, then appl
Command A+ sets a new high for Cohere's machine translation capabilities. Opening a clear gap over open source peers Mistral Medium 3.5, DeepSeek, & OpenAI's gpt-oss, as well as Claude Opus 4.6. A+ al
arXiv:2605.29001v1 Announce Type: cross Abstract: A paraphrase-quality audit of MathCheck (ICLR 2025) detected 4 semantically incorrect paraphrases in 129 groups (3.1%); removing them drops GPT-4o fro
arXiv:2605.30096v1 Announce Type: cross Abstract: Large language models (LLMs) can autonomously conduct multi-stage cyber attacks, but the consistency of their offensive behavior under repeated trials
arXiv:2605.28876v1 Announce Type: cross Abstract: CI failure logs are large (median 5k lines, max 200k in this corpus) and noisy. Coding agents that try to debug them depend on an upstream tool to red
OpenAI launches Rosalind Biodefense, expanding trusted access to GPT-Rosalind for vetted developers and U.S. government partners advancing biodefense, public health, and pandemic preparedness through
arXiv:2605.29744v1 Announce Type: new Abstract: The impressive performance of generalist large language models (LLMs) such as GPT and Claude in healthcare raises a critical question: will domain-speci
Worked on some code this morning using Opus 4.8 and so far I'm really liking it. Much more cooperative than 4.7 and less 'over agentic'. Stops and asks for my input when needed in places 4.7 (and GPT
I had Opus 4.8 in Claude Code write a sophisticated, if minor, academic paper from a archive of hundreds of de-identified research files from years ago I had to use GPT-5.5 Pro as a reviewer, it spott
Opus 4.8 formulated the hypotheses in advance, conducting data cleaning, did research on references, conducted analyses, did robustness checks, and put out the whole paper in LaTEX style. GPT-5.5 foun