Knowledge catalogue
arXiv:2605.03950v1 Announce Type: new Abstract: Although recent LMMs have become much stronger at visual perception, they remain unreliable on problems that require multi-step reasoning over visual ev
arXiv:2605.02620v1 Announce Type: new Abstract: Reproducing an empirical NLP study used to take weeks. Given the released data and a modern agentic-research harness, we redo every experiment of a rece
arXiv:2605.02348v1 Announce Type: new Abstract: Large language models pick up social biases from the data they are trained on and carry those biases into downstream applications, often reinforcing ste
arXiv:2605.02815v1 Announce Type: new Abstract: Text-to-SQL over large analytical databases requires navigating complex schemas, resolving ambiguous queries, and grounding decisions in actual data. Mo
arXiv:2605.01417v1 Announce Type: new Abstract: Evaluating large language models (LLMs) for medical applications remains challenging due to benchmark saturation, limited data accessibility, and insuff
arXiv:2605.02236v1 Announce Type: cross Abstract: Recursive language-model loops often settle into recognizable attractor-like patterns. The practical question is how much injected text is needed to m
arXiv:2605.01699v1 Announce Type: new Abstract: Recent attacks show that behavioural unlearning of large language models leaves internal traces recoverable by adversarial probes. We characterise where
arXiv:2605.00334v1 Announce Type: cross Abstract: Production agentic systems make many model calls per user request, and most of those calls are short, structured, and routine. This raises a practical
arXiv:2605.00677v1 Announce Type: new Abstract: While Large Language Models have achieved notable success on formal mathematics benchmarks such as MiniF2F, it remains unclear whether these results ste
arXiv:2604.27720v1 Announce Type: new Abstract: Deploying vision-language models (VLMs) in clinical settings demands auditable behavior under realistic failure conditions, yet the failure landscape of
arXiv:2604.27470v1 Announce Type: new Abstract: Millions of clinicians use ChatGPT to support clinical care, but evaluations of the most common use cases in model-clinician conversations are limited.
arXiv:2604.25067v1 Announce Type: cross Abstract: Forecasting when AI systems will become capable of meaningfully accelerating AI research is a central challenge for AI safety. Existing benchmarks mea
arXiv:2601.21225v2 Announce Type: replace Abstract: Large language models have made substantial progress in mathematical reasoning. However, benchmark development for multilingual evaluation has lagge
arXiv:2604.25102v1 Announce Type: new Abstract: Typographic prompt injection exploits vision language models' (VLMs) ability to read text rendered in images, posing a growing threat as VLMs power auto
We’re excited to introduce KAME: Tandem Architecture for Enhancing Knowledge in Real-Time Speech-to-Speech Conversational AI, accepted at #ICASSP2026! 🐢 Blog https://pub.sakana.ai/kame/ Paper https://
arXiv:2604.23091v1 Announce Type: new Abstract: Scaling EEG foundation models requires pooling data across heterogeneous electrode montages, a prerequisite both for larger pretraining corpora and for
Congrats to @poolsideai on Laguna XS.2 release. Day 0 support in @vllm_project. FP8, INT4, and NVFP4 quantized checkpoints created with LLM Compressor also ready today. 33B params, 3B active. Apache 2
arXiv:2604.23341v1 Announce Type: cross Abstract: The deployment of Large Language Models (LLMs) as assistants in electric grid operations promises to streamline compliance and decision-making but exp
arXiv:2603.15130v2 Announce Type: replace Abstract: Indirectness is a common feature of daily communication, yet is underexplored in NLP research for both low-resource as well as high-resource languag
arXiv:2604.24082v1 Announce Type: cross Abstract: Large (vision-)language models exhibit remarkable capability but remain highly susceptible to jailbreaking. Existing safety training approaches aim to
arXiv:2510.17590v2 Announce Type: replace Abstract: We present MERIT, an inference-time modular framework for multimodal misinformation detection that decomposes verification into four specialized mod
arXiv:2604.23530v1 Announce Type: cross Abstract: Multi-turn, long-horizon tasks are increasingly common for large language models (LLMs), but solving them typically requires many sequential model inv
arXiv:2505.16888v4 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed via third-party system prompts downloaded from public marketplaces. We identify a criti
Never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant to the user's query. — OpenAI Codex base_instruct
arXiv:2604.22778v1 Announce Type: cross Abstract: We present the first systematic study of weight matrix singular value spectra during transformer pretraining, tracking full SVD decompositions of ever
arXiv:2508.09603v2 Announce Type: replace Abstract: Membership inference attacks serves as useful tool for fair use of language models, such as detecting potential copyright infringement and auditing
arXiv:2604.24525v1 Announce Type: cross Abstract: Automated code review (ACR) bots are increasingly used in industrial software development to assist developers during pull request (PR) review. As ado
arXiv:2604.23434v1 Announce Type: cross Abstract: Dynamic Tanh (DyT) removes LayerNorm by bounding activations with a learned tanh(alpha x). We show that this bounding is a regime-dependent implicit r
Symphony is an open-source specification for orchestration released by OpenAI, designed to provide standardized patterns and tools for coordinating complex AI workflows and system interactions. The sp
arXiv:2604.22192v1 Announce Type: new Abstract: Chart-to-code generation demands strict visual precision and syntactic correctness from Vision-Language Models (VLMs). However, existing approaches are
Creative studios are being asked to do something fundamentally different now. Not just great work, but foundational systems for experience, brand, and content that their clients can own, extend, and b
Hermes will now by default use native vision if the main agent model supports it, and you didn't set a different vision auxiliary model! Just `hermes update` and it will take affect immediately. You c
arXiv:2510.07632v2 Announce Type: replace Abstract: Frontier AI models have achieved remarkable progress, yet recent studies suggest they struggle with compositional reasoning, often performing at or
arXiv:2508.06165v4 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have shown strong capabilities through two complementary paradigms: Retrieval-Augmented Generation (RAG) for know
arXiv:2604.22273v1 Announce Type: new Abstract: Iterative self-correction is widely used in agentic LLM systems, but when repeated refinement helps versus hurts remains unclear. We frame self-correcti
“AI: it feels most impressive when you don’t know much about the subject, don’t care or don’t have a clear idea of what the you want… Once you know what something should feel or look like, it becomes
extremely happy that we are in q2 2026, and engineers i look up to are plowing a path for the local model future. yes, we are still beholden to some lab publishing weights. but i take that over compan
Hermes Agent tip of the day: There are 4 ways to deal with the model while its running, - Message it, by default, it will interrupt the agent loop, stopping it and making it respond to your new messag
Sebastian Raschka has updated his LLM architecture gallery with higher resolution figures and improved summaries, including coverage of the DeepSeek V4 model architecture. This resource provides visua
If using a cloud based browser backend in Hermes Agent, it will now auto-detect if you want it to look at or use a locally hosted website and switch to local browser so it can access it. Hopefully thi
OpenAI's core principles emphasize democratizing access to AI technology, empowering individuals and organizations through AI tools, promoting universal prosperity across society, building resilient A
Sam Altman praised OpenAI's o1 model (version 5.5) for its earnest, persistent approach to problem-solving, comparing it to the 'little engine that could' mentality. The comment reflects Altman's pers
Bet this happens with Navier Stokes and it’s going to be something not even related to PDEs that solves it 23 years old with no advanced mathematics training solves Erdős problem with ChatGPT Pro. 'Wh
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
arXiv:2604.21611v1 Announce Type: cross Abstract: Inference-time scaling for LLM reasoning has focused on three axes: chain depth, sample breadth, and learned step-scorers (PRMs). We introduce a fourt
These pelicans are kind of angry looking! Left is deepseek-v4-flash, right is deepseek-v4-pro - both generated using OpenRouter via my LLM tool 🚀 DeepSeek-V4 Preview is officially live & open-sourced!
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
arXiv:2604.20791v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly deployed in healthcare, yet their communicative alignment with clinical standards remains insufficiently
Claude users can access more apps with Anthropic's AI now thanks to new connectors for everything from hiking to grocery shopping. Anthropic already supported connecting numerous work-related apps to
arXiv:2604.19754v1 Announce Type: new Abstract: Automated scoring of students' scientific explanations offers the potential for immediate, accurate feedback, yet class imbalance in rubric categories p
If you want to stack rank LLMs/VLMs on document understanding 📄, you can through ParseBench, now live on @kaggle 📊 ParseBench is the most comprehensive document OCR benchmark over real enterprise docu
arXiv:2604.19937v1 Announce Type: cross Abstract: Assessing chronic wound infection from photographs is challenging because visual appearance varies across wound etiologies, anatomical locations, and
Oh? Today we release a novel AI-assisted resolution of one of physics’ longest-standing questions. Given only: • Relativity as an axiom • One characteristic of the algebra A positive cosmological cons
Opus 4.7 just wrote a custom WebGPU kernel that runs Qwen3.5 up to 13x faster using a fused LinearAttention op! 🤯 Agentic kernel optimization is the future. Now live in 🤗 Transformers.js v4.2.0! P.S.
Thanks to @lmsysorg ！ Try it on SGLang now!🚀🚀 🚀 Qwen3.6-27B is here, and we have day 0 support on SGLang ✅ 27B params, beats Qwen3.5-397B-A17B across major coding benchmarks → Agentic coding → Text +
there are early signs of 5.5 being a competent ai research partner. several researchers let 5.5 run variations of experiments overnight given only a high level algorithmic idea, wake up to find comple
You can now use OpenAI OAuth to generate images with Hermes Agent! Update Hermes and use the ‘hermes tools’ command to configure the image generation tool to use OpenAI! Reminder: we have a $26,000 on
zero shot Kimi K2.6, go try it out its a good model sir! this is @Kimi_Moonshot running on @togethercompute, @opencode harness prompt below👇 Media Introducing Kimi K2.6 from @Kimi_Moonshot, a multimod
Also agree with this. AI's ability to do consistent pacing is a problem. Here, the poem anchors enough of the comic that it is okay, but static, though with a few good moments. But fiction writing and
An advantage of the humanities — of reading & seeing a wide range of great works from many perspectives & cultures — is that you develop your sense of taste. In a world where anyone can produce a floo