frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Start all of your commands with a comma (2009)

https://rhodesmill.org/brandon/2009/commands-with-comma/
232•theblazehen•2d ago•67 comments

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

https://openciv3.org/
694•klaussilveira•15h ago•206 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
6•AlexeyBrin•59m ago•0 comments

The Waymo World Model

https://waymo.com/blog/2026/02/the-waymo-world-model-a-new-frontier-for-autonomous-driving-simula...
962•xnx•20h ago•554 comments

How we made geo joins 400× faster with H3 indexes

https://floedb.ai/blog/how-we-made-geo-joins-400-faster-with-h3-indexes
130•matheusalmeida•2d ago•35 comments

Unseen Footage of Atari Battlezone Arcade Cabinet Production

https://arcadeblogger.com/2026/02/02/unseen-footage-of-atari-battlezone-cabinet-production/
67•videotopia•4d ago•6 comments

Vocal Guide – belt sing without killing yourself

https://jesperordrup.github.io/vocal-guide/
53•jesperordrup•5h ago•24 comments

Jeffrey Snover: "Welcome to the Room"

https://www.jsnover.com/blog/2026/02/01/welcome-to-the-room/
36•kaonwarb•3d ago•27 comments

ga68, the GNU Algol 68 Compiler – FOSDEM 2026 [video]

https://fosdem.org/2026/schedule/event/PEXRTN-ga68-intro/
10•matt_d•3d ago•2 comments

Show HN: Look Ma, No Linux: Shell, App Installer, Vi, Cc on ESP32-S3 / BreezyBox

https://github.com/valdanylchuk/breezydemo
236•isitcontent•15h ago•26 comments

Monty: A minimal, secure Python interpreter written in Rust for use by AI

https://github.com/pydantic/monty
233•dmpetrov•16h ago•124 comments

Where did all the starships go?

https://www.datawrapper.de/blog/science-fiction-decline
32•speckx•3d ago•21 comments

Show HN: I spent 4 years building a UI design tool with only the features I use

https://vecti.com
335•vecti•17h ago•147 comments

Hackers (1995) Animated Experience

https://hackers-1995.vercel.app/
502•todsacerdoti•23h ago•244 comments

Sheldon Brown's Bicycle Technical Info

https://www.sheldonbrown.com/
386•ostacke•21h ago•97 comments

Show HN: If you lose your memory, how to regain access to your computer?

https://eljojo.github.io/rememory/
300•eljojo•18h ago•186 comments

Microsoft open-sources LiteBox, a security-focused library OS

https://github.com/microsoft/litebox
361•aktau•22h ago•185 comments

UK infants ill after drinking contaminated baby formula of Nestle and Danone

https://www.bbc.com/news/articles/c931rxnwn3lo
8•__natty__•3h ago•0 comments

An Update on Heroku

https://www.heroku.com/blog/an-update-on-heroku/
424•lstoll•21h ago•282 comments

PC Floppy Copy Protection: Vault Prolok

https://martypc.blogspot.com/2024/09/pc-floppy-copy-protection-vault-prolok.html
68•kmm•5d ago•10 comments

Dark Alley Mathematics

https://blog.szczepan.org/blog/three-points/
96•quibono•4d ago•22 comments

Was Benoit Mandelbrot a hedgehog or a fox?

https://arxiv.org/abs/2602.01122
21•bikenaga•3d ago•11 comments

The AI boom is causing shortages everywhere else

https://www.washingtonpost.com/technology/2026/02/07/ai-spending-economy-shortages/
19•1vuio0pswjnm7•1h ago•5 comments

How to effectively write quality code with AI

https://heidenstedt.org/posts/2026/how-to-effectively-write-quality-code-with-ai/
264•i5heu•18h ago•216 comments

Delimited Continuations vs. Lwt for Threads

https://mirageos.org/blog/delimcc-vs-lwt
33•romes•4d ago•3 comments

Introducing the Developer Knowledge API and MCP Server

https://developers.googleblog.com/introducing-the-developer-knowledge-api-and-mcp-server/
64•gfortaine•13h ago•28 comments

I now assume that all ads on Apple news are scams

https://kirkville.com/i-now-assume-that-all-ads-on-apple-news-are-scams/
1076•cdrnsf•1d ago•460 comments

Female Asian Elephant Calf Born at the Smithsonian National Zoo

https://www.si.edu/newsdesk/releases/female-asian-elephant-calf-born-smithsonians-national-zoo-an...
39•gmays•10h ago•13 comments

Understanding Neural Network, Visually

https://visualrambling.space/neural-network/
298•surprisetalk•3d ago•44 comments

I spent 5 years in DevOps – Solutions engineering gave me what I was missing

https://infisical.com/blog/devops-to-solutions-engineering
154•vmatsiiako•20h ago•72 comments
Open in hackernews

Show HN: Terminal-Bench-RL: Training long-horizon terminal agents with RL

https://github.com/Danau5tin/terminal-bench-rl
125•Danau5tin•6mo ago
After training calculator agent via RL, I really wanted to go bigger! So I built RL infrastructure for training long-horizon terminal/coding agents that scales from 2x A100s to 32x H100s (~$1M worth of compute!) Without any training, my 32B agent hit #19 on Terminal-Bench leaderboard, beating Stanford's Terminus-Qwen3-235B-A22! With training... well, too expensive, but I bet the results would be good!

*What I did*:

- Created a Claude Code-inspired agent (system msg + tools)

- Built Docker-isolated GRPO training where each rollout gets its own container

- Developed a multi-agent synthetic data pipeline to generate & validate training data with Opus-4

- Implemented a hybrid reward signal of unit test verifiers & a behavioural LLM judge.

*Key results*:

- My untrained Qwen3-32B agent achieved 13.75% on Terminal-Bench (#19, beats Stanford's Qwen3-235B MoE)

- I tested training to work stably on 32x H100s distributed across 4 bare metal nodes

- I created a mini-eval framework for LLM-judge performance. Sonnet-4 won.

- ~£30-50k needed for full training run of 1000 epochs (I could only afford testing )

*Technical details*:

- The synthetic dataset ranges from easy to extremely hard tasks. An example hard task's prompt:

"I found this mystery program at `/app/program` and I'm completely stumped. It's a stripped binary, so I have no idea what it does or how to run it properly. The program seems to expect some specific input and then produces an output, but I can't figure out what kind of input it needs. Could you help me figure out what this program requires?"

- Simple config presets allow training to run on multiple hardware setups with minimal effort.

- GRPO used with 16 rollouts per task, up to 32k tokens per rollout.

- Agent uses XML/YAML format to structure tool calls

*More details*:

My Github repos open source it all (agent, data, code) and has way more technical details if you are interested!:

- Terminal Agent RL repo

- Multi-agent synthetic data pipeline repo

I thought I would share this because I believe long-horizon RL is going to change everybody's lives, and so I feel it is important (and super fun!) for us all to share knowledge around this area, and also have enjoy exploring what is possible.

Thanks for reading!

Dan

(Built using rLLM RL framework which was brilliant to work with, and evaluated and inspired by the great Terminal Bench benchmark)

Comments

rboyd•6mo ago
Great work! There should be a way for entities to crowdfund model training. Can a model like this be partially evaluated during training time and save through early stopping?

What are the best papers/resources on sota long-horizon RL?

Thanks.

thomasfromcdnjs•6mo ago
How much did you spend?
tjungblut•6mo ago
If you are curios, like me, how the actual reinforcement learning happens. It uses verl [1] underneath. The paper "HybridFlow: A Flexible and Efficient RLHF Framework" [2] explains it really well.

[1] https://github.com/volcengine/verl

[2] https://arxiv.org/abs/2409.19256v2

OtherShrezzing•6mo ago
That you've spent in the low-thousands (by the looks of it), and managed to beat GPT4.1 is an amazing insight into the moat of the big AI labs.
bravesoul2•6mo ago
Wow amazing! Amazing a "one person band" can do this much. It crosses many skillets.
erdaltoprak•6mo ago
This is incredible work
enigma101•6mo ago
Did you consider a kickstarter to overcome the gpu poorness??? 30 to 50 should be doable
anorwell•6mo ago
Some of the comments so far seem to be misunderstanding this submission. As I understand it:

1. Custom scaffolding (system prompt and tools) using Qwen3-32B achieved 13.75% on Terminal-Bench. No training was involved.

2. The author has built an RL system, but it has not been used for anything due to cost limitations.

So there's actually no result related to training here. It well known that the scaffolding used can have a large impact on benchmark outcomes (the Terminal bench leaderboard also demonstrates this [1]).

[1] https://www.tbench.ai/leaderboard

esafak•6mo ago
It looks like the submission has two aspects that are being conflated.

1. Tooling for training a terminal agent.

2. An agent that was _not_ trained with this tooling but prompt engineered. I could not find the author's discussion on this point.

TarasBob•6mo ago
I'm willing to help fund this if the creator is interested. I sent him an email.
lostmsu•6mo ago
Why do you need 50k? Can't you tune using LoRA?
Danau5tin•6mo ago
Exactly my first thought when I realised the cost! Currently LoRA is not supported by rLLM (The team told me they aim to support in next release), but it is certainly possible to port to verl directly or another RL framework for sure. I just did not have the time to port again (already done 2x as other RL frameworks had issues)