Knowledge catalogue
Recently I've flipped from being bullish to being bearish about AI. I think I'm updating my bearishness to be more solidly bearish. Early thoughts (which I hope to be disproven in the next year or so,
OpenAI field report: eight life-sciences projects used Codex/Claude Code agents; bottleneck shifts from coding to verification.
arXiv:2607.22798v1 Announce Type: cross Abstract: Computer-use agents are usually improved by strengthening perception: better models for reading a screenshot and choosing where to click. Yet a screen
arXiv:2607.23838v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG) lets a large language model answer questions using documents retrieved from an external knowledge base at query t
Grok Build with Grok 4.5 stands far ahead on the efficiency frontier....completely alone inside the chart’s most attractive quadrant It delivers top-tier coding-agent performance while using only arou
arXiv:2606.16794v2 Announce Type: replace Abstract: This study proposes a domain-specific LLM-based Visual Explanation Evaluation Framework for assessing visual attention explanations in facial skin d
I am writing a fairly complex C++ windows MFC application. I have a few 3090s and can run F16 Qwen3.6-27B with 256K context and MTP. The quality of code is exceptional with this quant vs its lower qua
I am setting up codex+ollama+qwen3.6:27b for my hobby coding project on a windows pc with rtx5090 - earlier i tried to setup vllm in Ubuntu container but couldn’t get that to work - now using ollama,
There are a few reasons why problems from International Mathematical Olympiad function as a good benchmark for LLMs: - The problems are new, not included in the training data of any model - Hard math
Hi everyone, I have a system which I have been using for gaming, R7 5700X, 48GB DDR4, RTX3090 24GB. But I want to use it for Local AI to reduce my reliance on cloud AI providers (mainly usage limits -
I am a long time iOS app developer. In the last year I have been using Cursor+Claude/others to assist with app development. I am concerned that the current low pricing will disappear eventually. I am
https://preview.redd.it/yw41s1jikefh1.png?width=1942&format=png&auto=webp&s=3a180ae6443c1db9f7b0ce621533a4b2aa553921 Hi, I've been running Ollama on my Unraid server since the llama2 era. I use to be
RTX 4070, 32gb system ram, Linux. NVIDIA-SMI 610.43.03, KMD Version: 610.43.03, CUDA UMD Version: 13.3 Systemd service modifications: [Service] Environment='OLLAMA_HOST=0.0.0.0:11434' Environment='OLL
Ruff v0.16.0 Astral shipped a significant new version of their Ruff Python linting tool a few days ago on July 23rd. I noticed today because my various CI jobs all started failing thanks to new defaul
The Opus 5 system card itself is a fun PDF to parse. It's 193 pages and stacked with labeled and unlabeled charts 📊 LlamaParse does a surprisingly good job on agentic (1.25c per page) and agentic plus
An interesting thing I'm observing from the blog/system card is that on a good chunk of the reported benchmarks (~20-30% from a skim), Opus 5 max thinking leads to a degradation in performance compare
arXiv:2607.20536v1 Announce Type: new Abstract: Tool-use agents that address day-to-day digital tasks such as ordering groceries must not only operate applications, but also interact with the user, e.
arXiv:2607.20526v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in settings where fluent but incorrect answers can be costly. In these settings, accuracy alone i
arXiv:2607.20645v1 Announce Type: cross Abstract: We introduce Frontier Financial Judgement, a challenging new benchmark developed in collaboration with professional equity analysts to assess agents'
arXiv:2607.20759v1 Announce Type: cross Abstract: AI coding agents powered by LLMs are increasingly integrated into real-world software development, where they generate, edit, and execute code with au
This is an open source and free tax engine which scored 96% on TaxCalcBench [highest ever recorded score till date] surpassing fable 5 and sol with just sonnet 5. The only 2 cases where it missed, it
Opus 5 is a great model for coding, data analysis, design, biology, knowledge work. More than any of these eval scores, what is most exciting to me is something else: Opus 5 is our least prompt inject
arXiv:2607.20864v1 Announce Type: cross Abstract: Position bias in multiple-choice LLM evaluation is widely cited as a confound in capability comparisons, but published measurements rely on single ans
arXiv:2607.20911v1 Announce Type: new Abstract: We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring pro
Announcing OpenWorker! An open-source agent that doesn't just chat with you, but delivers finished work -- like hand you a polished document, send a slack message, or update a calendar entry. Ask it t
I wanted to know how cheap you can go and still run local models, so I ran Ollama CPU-only on a Youyeetoo X1S. It's a single-board x86 machine with a Celeron N5095 (Jasper Lake, 4C/4T, 15W), 16GB of R
arXiv:2607.19375v1 Announce Type: cross Abstract: Language models perform economically valuable work, yet they are not currently assessed for how well they perform every economically valuable task. We
The interesting finding from a new [arXiv paper](https://arxiv.org/abs/2607.16165) isn't that a frontier vision model failed a new benchmark, that happens weekly, but the specific shape of the failure
If you are building real-time voice agents with @GoogleDeepMind Gemini Live, you can now trace your speech-to-speech agent loops directly in @LangChain! - Speaker callback hooks capture only the exact
The first router built for generative media is here. Runway launches AI model router as generative media gets crowded https://techcrunch.com/2026/07/23/runway-bets-on-ai-model-routing-as-generative-me
arXiv:2607.19967v1 Announce Type: cross Abstract: Shippers are beginning to delegate carrier selection to large language model (LLM) agents. We ask what such delegation does to a freight matching mark
Anthropic PBC will purchase up to 2 gigawatts’ worth of graphics cards from Advanced Micro Devices Inc. as part of a multibillion-dollar deal announced today. The partnership also has several other co
Are AI labs pelicanmaxxing? Excellent piece of work by Dylan Castillo, who took a deep-dive into the frequently pondered question of whether the AI labs have been deliberately training models to draw
webgpu : add CONV_2D_DW (depthwise conv2d) kernel (#25847) webgpu : add CONV_2D_DW (depthwise conv2d) kernel Implement GGML_OP_CONV_2D_DW for the WebGPU backend, ported from the Vulkan backend's conv2
“Generate a fake, but believable, witty Churchill insult at a party and explain the context. It should be very clever and original” This time, I think GPT 5.6 Sol Pro wins, but Fable is good too, and
One thing I noticed in American politics, whenever the government wants to push unpopular actions or laws, they often introduce fear to convince the public to support them. This is actually how i view
Hey everyone, I’m a Full-Stack Developer with 6+ years of experience. I’m relatively new to AI-assisted development workflows and want to build a production-ready, enterprise-level Next.js web applica
This is OpenSWE! It's an OSS coding agent that runs in the cloud and lives in Slack. Use it for coding, general Q&A, planning, etc etc etc. It's truly a jack of all trades (and very widely used at Lan
BREAKING: Kimi K3 by @Kimi_Moonshot is 1st overall on 3D Design with an Elo of 1450. This is a 6 position and 108 Elo jump from @Kimi_Moonshot's previous model, Kimi K2.6. This performance puts Kimi K
LWiAI Podcast #249 (June 17 2026) reports that Anthropic halted access to Fable 5 and Mythos 5 following a U.S. government order over alleged jailbreaks, prompting debate on export controls and the fe
Huge launch from @tryramp. Different steps in an agent workflow can use different models. This can help reduce costs significantly without sacrificing performance. Model routing will become a core par
Open ecosystem or walled garden? For @anthropicai's @katelyn_lesse and @angjiang the answer is clear. 'We actually aren't precious about “You should run these things on our infrastructure.' In practic
Automated lint: 20 errors, 8743 warnings, 3 info
arXiv:2607.13071v1 Announce Type: cross Abstract: Agentic LLM coding tools compress long session histories into compaction summaries that subsequent sessions inherit as ground truth. This paper docume
Tool: Mermaid to Unicode box art (grok-mermaid) While exploring the codebase for the newly open-sourced Grok CLI coding agent I came across xai-grok-markdown/src/mermaid.rs, a 'self-contained terminal
arXiv:2601.02023v2 Announce Type: replace-cross Abstract: As Large Language Models (LLMs) increasingly utilize massive context windows as working memory for autonomous tasks, their reliability fluctua
arXiv:2607.10057v1 Announce Type: cross Abstract: Can AI agents visually comprehend quantum circuit diagrams and generate verified executable code--and at what cost? We present Quantum Circuit Vision,
arXiv:2607.13049v1 Announce Type: new Abstract: Foundation models have given robots a sophisticated brain for complex decision-making, yet deploying that intelligence into a physical platform still de
arXiv:2607.13618v1 Announce Type: new Abstract: LLM agents are increasingly evaluated on multi-week decision tasks in which the state that drives cost is never directly observed. On such tasks the fin
arXiv:2607.13346v1 Announce Type: cross Abstract: Alignment faking is dangerous because a model can appear compliant under monitoring while preserving behavior it would reveal when unmonitored. When n
arXiv:2607.12051v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly being explored for clinical decision support, but their reliability in complex oncology treatment planning
arXiv:2607.12195v1 Announce Type: cross Abstract: Semantic memory retrieval can be conceptualized as navigation through conceptual space. We compared semantic search dynamics between humans and three
arXiv:2607.12884v1 Announce Type: new Abstract: Patients seeking medical information often ask questions that embed incorrect assumptions or misconceptions. In such cases, safe medical communication r
I demonstrated the incredible power of o1-preview/ reasoning less than two years ago by showing it could solve this crossword puzzle with only one hint. https://www.oneusefulthing.org/p/something-new-
I’ve found @angjiang and @katelyn_lesse of @AnthropicAI to be generous, transparent, and thoughtful when it comes to building an ecosystem, not a walled garden. Listen and decide for yourself. Just so
just ran this on my phone (17 pro) Every on-device AI benchmark you've seen was measured on someone else's phone. So we stopped posting numbers and shipped the benchmark instead. Built into the RunAny
Automated lint: 26 errors, 6728 warnings, 3 info
Or… what if we gave you $100 in Codex credits if you tell us what you love about GPT-5.6 Sol or why you switched? Tweet it, claim your gift, enjoy more usage. First 10k get the free tokens! https://sw
arXiv:2607.12216v1 Announce Type: cross Abstract: Multi-agent and memory-augmented LLM systems often place coordination content, shared state, prior discussion, tool outputs, summaries, and role instr
arXiv:2607.12227v1 Announce Type: new Abstract: We revisit the evaluation of automatic harness evolution for LLM agents. Existing harness evolution methods use unit test cases to search for harness co