Ameesh Shah (@ameeshsh) header
Ameesh Shah (@ameeshsh) profile picture

Research Scientist at Waymo working on improving autonomous behavior. Prev: Ph.D @UCBerkeley, UG @RiceUniversity | he/him

CLE ➡️HTX ➡️Cal➡️SF
Joined April 2012
Photo by Ameesh Shah (@ameeshsh)Photo by Ameesh Shah (@ameeshsh)Photo by Ameesh Shah (@ameeshsh)Photo by Ameesh Shah (@ameeshsh)Photo by Ameesh Shah (@ameeshsh)Photo by Ameesh Shah (@ameeshsh)
Ameesh Shah retweeted
How can robot policies be trained to best leverage VLMs' CoT reasoning and in-context learning for generalization? The key is Steerable Policies: vision-language-action models that can be flexibly controlled in many ways! steerable-policies.github.io 1/9
6
34
7
138
24,271
Ameesh Shah retweeted
Flow policies are getting popular in robotics as they capture multimodal prior data well, and synergize nicely with action chunking. But it is unclear how to best train them with RL effectively. We found something that works pretty well! (spoiler: use Adjoint Matching) 🧵1/N
12
55
3
458
90,165
🗣️📣Announcing VerifAI 2: AI Verification in the Wild, an upcoming workshop at #ICLR2026!! 🗣️📣 VerifAI will gather researchers to explore topics at the intersection of genAI and trustworthy ML. Submit your work! Check out our website and CFP for more: verifai-workshop.github.io/
7
2
25
10,064
Action chunking is drawing growing interest in RL, yet its theoretical properties are still understudied. We are excited to share some insights on when we should use action chunking in Q-learning + a new algo (DQC) to tackle hard long-horizon tasks!colinqiyangli.github.io/dqc🧵1/N
6
56
4
307
90,159
I’m at #NeurIPS2025 until Sunday! Reach out to chat about multi-agent learning, AI safety, human-AI interaction, or LM agents. Excited to present our work on rational policy gradient in hall C,D,E, #5511, 11-2pm on Wed with @MichaelD1729 @MicahCarroll @ameeshsh
Our NeurIPS 2025 paper extends adversarial learning (adversarial examples, self-play, etc.) beyond zero-sum games by solving "self-sabotage". 🧵👇
1
7
2
33
5,300
Had a lot of fun working on this project with @NiklasLauffer et. al. 🤠 We're able to improve multi-agent policies by having agents play with *meaningfully challenging* partners! But finding those partners is not an easy task - check out the thread for more!
Our NeurIPS 2025 paper extends adversarial learning (adversarial examples, self-play, etc.) beyond zero-sum games by solving "self-sabotage". 🧵👇
2
12
1,391
Ameesh Shah retweeted
Now we can have self-refinement for VLAs in the real world (with the aid of a big VLM)! VLM critiques VLA rollouts and iteratively refines the commands to make it perform better.
LLMs have shown a remarkable ability to “self-refine” and learn from their mistakes via in-context learning. But in robotics, most methods are single-shot. How can we bring inference-time adaptation to robot learning? A 🧵:
5
21
242
27,154
LLMs have shown a remarkable ability to “self-refine” and learn from their mistakes via in-context learning. But in robotics, most methods are single-shot. How can we bring inference-time adaptation to robot learning? A 🧵:
10
19
3
131
38,164
As a result, LITEN’s VLM planner can iteratively learn both the VLA’s capabilities and relevant information about real-world dynamics to solve complex, long-horizon tasks.
1
1
7
869
Everyone knows action chunking is great for imitation learning. It turns out that we can extend its success to RL to better leverage prior data for improved exploration and online sample efficiency! colinqiyangli.github.io/qc/ The recipe to achieve this is incredibly simple. 🧵 1/N
3
76
5
366
48,692
The submission deadline to our workshop has been extended until Feb. 7!!!! Submit your papers on trustworthiness + verification + genAI + ML and come hang with us in Singapore!
📣Announcing VerifAI: AI Verification in the Wild, a workshop at #ICLR2025 VerifAI will gather researchers to explore topics at the intersection of genAI/trustworthyML and verification: verifai-workshop.github.io/ @celine_ylee @theo_olausson @ameeshsh @wellecks @taoyds
4
6
2,115
The ultimate test of any physics simulator is its ability to deliver real-world results. With MuJoCo Playground, we’ve combined the very best: MuJoCo’s rich and thriving ecosystem, massively parallel GPU-accelerated simulation, and real-world results across a diverse range of robot platforms: quadrupeds, humanoids, dexterous hands, and arms. Best of all? You can get started today with a single command: pip install playground playground.mujoco.org/
37
177
41
883
154,369
Ameesh Shah retweeted
📣Announcing VerifAI: AI Verification in the Wild, a workshop at #ICLR2025 VerifAI will gather researchers to explore topics at the intersection of genAI/trustworthyML and verification: verifai-workshop.github.io/ @celine_ylee @theo_olausson @ameeshsh @wellecks @taoyds
21
3
82
30,360
Ameesh Shah retweeted
📢📢 Thrilled to share our new work: Syzygy: Dual Code-Test C to Rust Translation using LLMs and Dynamic Analysis! Key Principles 1️⃣ 🤖🤝🔍 Combining LLM Inference Scaling with Dynamic Analysis: * Best of both worlds (program semantics and neural search) 2️⃣ ⚡️🧪 Dual Code-Test Translation: * Don't just convert the code; translate tests for incremental validation Results: We successfully translated Zopfli, a C compression library with ~3000 LoC to Safe Rust with test-based validation 🤯 1/
1
20
3
70
13,406