Links: 📄 Paper: arxiv.org/abs/2606.08340 🌐 Website: alem-world.github.io 💻 Env + harness: github.com/alem-world/a... 🤗 MARL baselines: huggingface.co/alem-world/a... Try your own agent and submit it!
Who's Who of AI
Kale-ab Tessera
Why they matter
Researcher with public evidence across AI research, Agents & robotics.
- AI signals
- 5 past 30d
- Sources
- 4 distinct domains
- Discussions
- 1 past 30d
- Latest signal
- 16d ago
ML PhD Student @ Uni. of Edinburgh, working on Multi-Agent Problems. | Organiser @deeplearningindaba.bsky.social @rl-agents-rg.bsky.social | 🇪🇹🇿🇦
kaleabtessera.com
What they're sharing
Benchmarking Open-Ended Multi-Agent Coordination in Language Agents arxiv.org
GitHub - alem-world/alem-env: Can LLM agents coordinate in long-horizon, open-ended tasks? github.com
alem-world/alem-rl-baselines · Hugging Face huggingface.co
Probing Dec-POMDP Reasoning in Cooperative MARL arxiv.org
Alem: Open-Ended Multi-Agent Coordination alem-world.github.io
Alem: Full interactive reasoning traces alem-world.github.io
Articles & links
Links: 📄 Paper: arxiv.org/abs/2606.08340 🌐 Website: alem-world.github.io 💻 Env + harness: github.com/alem-world/a... 🤗 MARL baselines: huggingface.co/alem-world/a... Try your own agent and submit it!
Links: 📄 Paper: arxiv.org/abs/2606.08340 🌐 Website: alem-world.github.io 💻 Env + harness: github.com/alem-world/a... 🤗 MARL baselines: huggingface.co/alem-world/a... Try your own agent and submit it!
We open-sourced the library for computing the metrics and have a fun website you can play around with to get the intuition. 📜: arxiv.org/abs/2602.20804 💻: kaleabtessera.com/probing 🧑💻: github.com/KaleabTesser...
Links: 📄 Paper: arxiv.org/abs/2606.08340 🌐 Website: alem-world.github.io 💻 Env + harness: github.com/alem-world/a... 🤗 MARL baselines: huggingface.co/alem-world/a... Try your own agent and submit it!
There's much more in the paper and website, including coordination by task type, model-size effects, and sampled interactive traces from all 13 models. The agent conversations alone are worth exploring, from role assignment to coordinated revives: alem-world.github.io/traces.html
Their own posts
Recent commentary
Can LLM agents coordinate in long-horizon, open-ended worlds? We test 13 LLMs in Alem, a new benchmark. Most struggle, averaging ~6% normalised return. Yet on Hard, zero-shot Gemini 3.1 Pro matches the best MARL agent after 1B training steps. Our ablations show communication matters most. 🧵
Their network
In Kale-ab Tessera's orbit
Center = Kale-ab Tessera. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.