Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
Web-agent benchmarks overwhelmingly measure depth -- pinning one obscure answer behind a chain of constraints -- while breadth, exhaustively enumerating a closed set and filling each…
Reinforcement learning (RL) post-training improves the reward alignment of flow-based generators, but often degrades perceptual quality in ways that are not captured by the reward proxy.…
Vision-language models (VLMs) are increasingly deployed in consumer, medical, financial, and enterprise applications. This broad deployment expands the safety surface: risks can arise from multimodal question…
RT AhmadWe actually kickstart things early tomorrow with 2 workshops on Local AISee you guys tomorrowAhmad: MASSIVE NEWSTeamed up with NVIDIA to make Local AI The…
Dean has sizable cajones himself if his poasting game is any indication; but this response to @zetalyrae is weak. Let's start with the fact: "we" (not…
> No cars were built in America in 1942, '43, '44, '45> We did build 144 aircraft carriers, 29,000 B-29sDo you feel it anon?? An HBM…
RT egeToday I was walking around in Sausalito and asking @grok voice how much apartments around me cost to rent and it was truly an AGI-ly…
RT SpaceXFalcon 9’s first stage has landed on the A Shortfall of Gravitas droneship
Like some demigod chilling in retirementBruce Thompson51: @woke8yearold Bruce Sterling still posts here occasionally I think, but rarely about the cyberpunk genre he helped found.
Interestingarctotherium: Did Elon Musk kill millions of people by cutting USAID? We now have preliminary 2025 mortality data from several African countries, and the answer may…
Extremely troublingiamyesyouareno: “We will kill the white man’s children, we will kill their woman, we will kill anything that we find in our way!” —Andile Mngxitama,…
RT SpaceXWatch Falcon 9 launch the @SiriusXM SXM-11 mission to orbit https://x.com/i/broadcasts/1AJEmmRWBjkJL
We present Qwen-Image-2.0-RL, a post-training pipeline that applies reinforcement learning from human feedback (RLHF) and on-policy distillation (OPD) to improve both the visual quality and instruction-following…
We introduce an axiomatic evaluation framework for latent thought representations in LLMs, comprising metrics that are independent of downstream benchmark scores and reveal representational failures that…
I know people consider llama 405b and others to be old now, lol, but I'm wondering if there would be a use case for it. I…
Artificial intelligence is driving a revolution in scientific discovery, accelerating everything from hypothesis generation to mathematical theorem proving. However, this rapid acceleration is creating a systemic…
Training and evaluating robot policies in the real world is costly and difficult to scale. We introduce SimFoundry, a modular and automated system for zero-shot real-to-sim…
Video generation models have emerged as a promising paradigm for embodied world simulation. However, both general-domain video generators and robot-specific data fine-tuned models can still produce…
> Trend extrapolation is your friend. Your best friend. Don't let anyone tell you otherwise. I've actually only rarely see someone extrapolate a trend too credulously;…
Article URL: https://act.eff.org/action/tell-congress-don-t-force-age-checks-online Comments URL: https://news.ycombinator.com/item?id=48713887 Points: 18 # Comments: 1
There's not much to this. Cyberpunk, like (almost?) every genre, adopts the POV of outlier individuals in a world with supernatural power scaling. A normie/goon NPC…
The CUDA tool kit in apt is waaaaaaaay too old. I’ve been struggling with compute issues after all the other tweaking. Ubuntu latest is 12.0 newest…
Article URL: https://danunparsed.com/p/hackerrank-open-source-ats Comments URL: https://news.ycombinator.com/item?id=48713832 Points: 14 # Comments: 0
I keep my recipies on my website, and like most of my website it grew over time instead of being designed. A couple years ago I…