Algorithmic Research Group
@algoresearch_
Our mission is to develop advanced AI models and agents designed to accelerate progress and unlock new discoveries in science.
Our ML Wiki is now editable! We're excited to work with the community on improving ground truth. Propose an edit to the wiki - it's sent to an admin for approval. You're notified when it's live!
We're adding new features daily. Check out this side-quest we went on - a wiki page for every CS paper on ArXiv, complete with a summary, models used, datasets used, methods, keywords, related papers, whether or not they claimed SOTA etc etc.
We build ScoutML for AI researchers (and agents) to simplify the lit review process. It's pretty easy to get started with the ScoutML API. Grab a key and start running some commands. Link below:
A key step towards automated AI R&D just landed on the @huggingface Hub today: > Step 1: Use an agent to pick the most promising (model, dataset) pair on the Hub via hf-mcp-server > Step 2: Post-train the model with TRL/@unsloth /@axolotl_ai and `hf jobs` > Step 3: Evaluate…
Very excited to launch this little tool that we’ve been building. ScoutML is an API built for AI researchers and agents that includes a ton of metadata on each paper. It’s been super helpful for us as we run our research agents internally. Hopefully it will help fill in some of…
At ARG, we're laser-focused on understanding recursive self-improvement. We're confident that as models scale, RSI will accelerate the frontier of AI at ever-increasing speeds. Over the past year, we've created benchmarks, agents, and AI systems to measure how this might happen.…
We're opening up our beta a little bit an would love more input! 📊
A lot of ML tools help you implement. Not many help you think. When I’m exploring a new research direction, I don’t want another search engine or citation graph. I want something that’s actually read the literature, can suggest promising directions, and helps me reason through…
🚨 AI/ML PhDs: Get paid to write the research idea you wish existed. 💡 Task: Propose a novel ML idea 💰 $500 for submission 🏆 $1K bonus for top 5 🎓 Ideal for PhDs, grads, profs & senior AI folks 🌍 Remote + async ⏱ ~4–6 hours Apply 👉 forms.gle/rdd3oUez2U6Au3…
Cool work. We've had this in our agents at @algoresearch_ for the past year but haven't stopped to measure it's performance formally. It clearly helps a lot
New research from our team at @AnthropicAI shows how giving Claude a simple 'think' tool dramatically improves instruction adherence and multi-step problem solving for agents. We've documented our findings in a blog post:
A few days ago, we open sourced our v0 agent for AI research and develop. Up next we'll be open-sourcing our v0.1 multiagent system, including agents for research, training, evaluation, and decision making. Here's v0 training a simple mlp on mnist - more examples upcoming
imho it is super important that we get numbers on agentic AI R&D now, while it's still in an early state, and that *both!* the benchmarks and the agents are open source. benchmark results on closed source agents add little value here
United States Trends
- 1. #WWERaw 55.2K posts
- 2. John Cena 55K posts
- 3. Packers 42.7K posts
- 4. Eagles 68.4K posts
- 5. #RawOnNetflix 1,424 posts
- 6. GRAND SLAM CENA 14.9K posts
- 7. #GoPackGo 4,523 posts
- 8. Jalen 15.1K posts
- 9. Jordan Love 3,364 posts
- 10. Tush Push 10.1K posts
- 11. Cade Horton 1,125 posts
- 12. Shipley N/A
- 13. Drake Baldwin 5,234 posts
- 14. Dirty Dom 1,583 posts
- 15. Green Bay 10.6K posts
- 16. Go Birds 17.5K posts
- 17. Monday Night Football 6,304 posts
- 18. Triple H 5,497 posts
- 19. Dominik Mysterio 10.9K posts
- 20. #ANDNEW 3,091 posts
Something went wrong.
Something went wrong.