When can we say AI made a scientific discovery?
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last Wednesday, Anthropic announced that earlier this year it
Knowledge catalogue
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last Wednesday, Anthropic announced that earlier this year it
Also kudos to fireworks AI for admitting that they built on top of Kimi (I remember the awkward Mistral 3 release) https://x.com/rasbt/status/1999557830131494977?s=20 Hold on a sec, Mistral 3 Large us
Elliot Arledge released a video titled “Claude Opus 5.5 just one‑shot,” which explains how to optimize CUDA kernels using a music‑video format. The clip was posted on September 27, 2026 and has accumu
Cool paper on catching alignment failures with Jev. The ideas is to ask Jev one generic yes/no question about a model's response, and use its probability as a score. With no extra training, that score
Early this year when I was first looking at building up my inference capability you could get the 16GB Tesla P100s for between 60 and 80. Asked claude about it, told me absolutely not worth it. No ten
Slopus is a desktop app for making AI videos on your own GPU, from the first prompt to the exported file. What it does: Generate scenes shot by shot. You lay out scenes on a board, write each shot, an
Where to even begin on this one...I guess I should start by acknowledging the risk vs reward for vendors other than nvidia, so: Yes I understand that nvidia are dominant currently on speed (llm and im
I built this, so the disclosure goes up front. It's a free, public MCP server plus a REST API with Canadian privacy law data. The MCP endpoint is at https://movahedi.ca/mcp and it uses Streamable HTTP
SpaceX's CFO says Starlink demand is unprecedented. The interesting part is who's driving it! Larry Goldberg @TeslaLarry breaks down the Goldman conference interview. Bret Johnsen reiterated a coming
I’ve spent the last couple of days with Opus 5.5 working on a fork of Inco’s excellent and already blazingly fast Splash engine to optimise it for M5 Max chips. Taking liberties and referring to it as
an implicit concession that they cannot control their models. Notable new disclosures from OpenAI that shouldn't get lost in all the other news. Of particular interest (and seems sensible!): '~all inf
Micha wrote “one of our models was able to gain unauthorized access to the internet during RL training” I love the use of the passive voice. Tis a subtle way to say “we humans at @OpenAI are essential
People on X: “CLAUDE IS BACK, baby!” Listen, Opus 5.5 is incredible and probably my favorite overall model right now, but it is still ridiculously verbose and over-justifies and over-qualifies everyth
Madison Mills / Axios: Sources: OpenAI, Anthropic, and researchers are probing tens of thousands of frontier model security incidents, including sandbox escapes and website hijacking — OpenAI, Anthrop
We comprehensively evaluated 16 recent frontier VLMs - including Opus 5.5 and GPT-6 Sol/Luna - on whether higher effort led to better document parsing performance. Higher effort typically leads to imp
Hi, human here with rambling thoughts to share. Feel free to skip Overall, I don’t feel like I’m missing much; if anything. On my hardware(M1 ultra w 128Gb) it’s probably not as fast as Claude but I’v
After the Hugging Face incident, we committed to conducting a much broader review of actions taken by our models during training and evaluation and to being transparent about our findings. This is an
arXiv:2609.29744v1 Announce Type: cross Abstract: We present: (i) a new dataset consisting of the full development history of a 21,000-line Python tool built entirely by Claude AI, with no human-autho
Check out this week's updates and releases: — Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two of our most expressive audio generation models yet — Gemini 3.8 Live with Live Avatar, bringing ne
Deploy the publicly available Qwen3-TTS-12Hz-1.7B-Base text-to-speech model from Amazon SageMaker JumpStart to a fully managed, real-time endpoint, and clone a voice from a short reference clip. Cross
arXiv:2609.29669v1 Announce Type: cross Abstract: Robot learning now advances along two tracks that rarely meet. On one side, direct Vision-Language-Action (VLA) policies map observations to actions a
Everyone thinks Claude is just good at writing code and hacking critical infrastructure. Few people know that Claude is also an accomplished artist, filmmaker, editor, cinematographer and art director
arXiv:2609.29692v1 Announce Type: new Abstract: Fair allocation of scarce, indivisible resources is an important challenge in many societal problems. While there are several formal theories of fairnes
How do you actually build an effective harness with Claude? We had Thariq (@trq212) from Anthropic at Navigate 2026 to talk about 'Unhobbling Claude', the difficulties and processes on how to build ag
arXiv:2606.14777v2 Announce Type: replace-cross Abstract: Many moments in the real world do not wait for a user to ask. A fire starts on a security monitor, an expression flickers across a video call,
I finally got around to upgrade the GPU's. First a word of warning, when upgrading GPU's on P520 you have to be extra careful not to slot the card on any angle other than straight when installing or w
arXiv:2609.30130v1 Announce Type: cross Abstract: Current vision-language models (VLMs) excel at visual content understanding and text-based reasoning, yet their structure limits the advancement of in
I’ve been experimenting with whether Qwen3.8-Flash-Next’s pretrained PLE n-gram memory can improve a much smaller Qwen3.5-0.8B model. I trained the 0.8B setup with limited resources, mostly using free
arXiv:2609.30219v1 Announce Type: cross Abstract: An acceptance protocol is developed for sensor-coordinate and polarity binding in mechatronic commissioning. Candidate generation is separated from re
arXiv:2609.29490v1 Announce Type: cross Abstract: Recent advances in robotics highlight hierarchical configurations of robot morphology, where multiple levels of functional substructures synergize to
arXiv:2609.28969v1 Announce Type: new Abstract: While end-to-end autonomous driving systems show promise, their application to micromobility vehicles is hindered by simulators failing to capture speci
TL;DR - Swift Flash is a killer model that massively reduces excess reasoning. Try it out! If you haven't seen from my previous comparison posts, I'm a huge fan of the Swift Qwen3.8 models. I've been
arXiv:2609.29423v1 Announce Type: new Abstract: No two robots are truly identical: calibration, battery state, sensor drift and wear give every swarm a distribution of behaviour rather than a single p
Hi people, I am looking for a lightweight IDE or plugin that won't inject large context at initiation. I tried Cline and native VS Code but they inject such heavy initial context that it fills up my g
Banger paper introducing Jev-as-a-Judge. The overall finding is that you want to use a cheap judge for most of your evals and send only the uncertain calls to a frontier model. This paper measures how
arXiv:2505.10083v2 Announce Type: replace Abstract: Conventional forecasting methods are trained end-to-end on unimodal time series, which limits their ability to exploit textual information and under
Robert Hart / The Verge: Experts say that air-gapping AI could prevent events like the Hugging Face hack, but would undermine the value of evaluations and slow research to a crawl — Researchers are te
If you believe the risk is coming from 1-3 people in a garage with no money and no compute, you just don't understand this technology and haven't learned anything this summer. Risk comes from the asym
Hi, sorry if this doesn't belong here. Getting to the point basically, I've been using online-only AI like GPT/Gemini since 2022, and have been interested in local models but am clueless overall. Yes
arXiv:2609.28364v1 Announce Type: cross Abstract: Control barrier functions (CBF) are a popular safety filter to ensure safety for nonlinear dynamical systems. However, when the system is subject to u
arXiv:2609.26872v1 Announce Type: new Abstract: Musculoskeletal (MSK) humanoids provide a physiologically grounded embodiment for studying full-body motor control, but their high-dimensional muscle ac
Guys I have a M4-mac air with 24 GB unified memory and I wanna run some local LLM model on my machine Currently I am working on building and RAG application which is mostly and QnA system but I wanna
arXiv:2609.28049v1 Announce Type: cross Abstract: Video understanding is usually benchmarked on curated, single-actor, or professionally filmed clips, and a strong score there is routinely read as evi
arXiv:2609.27463v1 Announce Type: new Abstract: Distributed model predictive control (DMPC) often constructs both predictions and collision constraints from neighbor trajectories, so packet loss can r
arXiv:2609.27175v1 Announce Type: cross Abstract: Multimedia verification requires not only accurate decisions but also traceable evidence, reliable human correction, and safe reuse of prior experienc
arXiv:2609.26989v1 Announce Type: new Abstract: Hexapod robots can achieve static stability with fewer actuated joints than bipeds or quadrupeds, yet many platforms still use 3-DOF legs, increasing we
Priyanka Salve / CNBC: VFX company DNEG's AI unit Brahma, whose customers include Warner Bros. and the NBA, raised 150M led by Indian PE firm Multiples at a 2B post-money valuation — Brahma AI, an Ind
arXiv:2609.28429v1 Announce Type: new Abstract: An always-on robot faces an endless stream that never resets: instructions arrive and lapse, the scene changes, and its own past actions reshape what it
arXiv:2609.27199v1 Announce Type: new Abstract: Sparse communication in decentralized zeroth-order learning requires compatible peer-state coordinates. We characterize this one-hop condition and devel
arXiv:2609.25470v1 Announce Type: new Abstract: Bioinspired feedback controls for pursuit, tracking, and collective motion are often expressed in terms of the relative configuration between interactin
arXiv:2602.01163v2 Announce Type: replace Abstract: Safe UAV emergency landing requires more than just identifying flat terrain; it demands understanding complex semantic risks (e.g., crowds, temporar
Robert Hart / The Verge: Anthropic says Claude autonomously discovered a new enzyme system in the DNA of bacteriophages, somewhat similar to CRISPR, the first result from its new biolab — It took near
arXiv:2609.24921v1 Announce Type: new Abstract: Scientific weak signals are early, low-visibility research directions that later become central to mature scientific topics, yet existing resources such
Can you take the cheapest open model from a frontier lab & beat closed models on the hardest long range tasks? Yep Our open adaptive intelligence system Zenith ramps DeepSeek v4.1 Flash past GPT 5.6 S
arXiv:2511.17909v2 Announce Type: replace Abstract: Chemical reasoning inherently integrates visual, textual, and symbolic modalities, yet existing benchmarks rarely capture this complexity, often rel
arXiv:2609.23064v1 Announce Type: new Abstract: Understanding the physical world requires more than object recognition, scene description, and short-term visual prediction, as real-world physical syst
arXiv:2609.23130v1 Announce Type: new Abstract: Large language model (LLM) inference is evolving from an engine-local optimization problem into a distributed control problem involving reusable state,
arXiv:2609.25363v1 Announce Type: new Abstract: Object transportation is a fundamental capability for humanoid robots operating in real-world, human-centric environments, yet existing methods struggle
arXiv:2609.22529v1 Announce Type: new Abstract: International law provides the normative framework through which states coordinate action, regulate armed conflict, and protect human rights, yet its te
I’m the developer of TensorSharp. Inspired by vLLM PR #57250, I added and optimized a native /v1/systemone endpoint for Jev-style structured decisions using DiffusionGemma GGUF. This uses DiffusionGem