Prithwish Jana

PhD (CS) Candidate @GeorgiaTech | Neuro-Symbolic AI, Self-Improving Agents & AI4Math | 2x Applied Sc. Intern @AWS | @IITKgp Silver Medalist | JU Gold Medalist

[Prithwish Jana @jprithwish]

[1/n]
🧬 Want a stronger agent without retraining? Give it a better harness.

MILO, our recent work from AWS AI Labs, does this automatically: +12.0% resolution rate on Terminal-Bench 2.1, and new records on 3 open math problems, beating AlphaEvolve, TTT-Discover and EvoX.

[Xinyu (Rachel) Li @RLixinyu]

🚀 We gave a 4B model with an 8K context window more test-time compute. Instead of improving, it got worse.

Scaling across context windows requires more than inline reasoning. It takes contextual reasoning: knowing how to delegate, preserve reasoning in compact summaries, and continue reasoning from them.
(1/10)

[Langlin Huang @shrangoh]

🧵 Can your teacher go off-policy in On-Policy Distillation?
Yes! An unfamiliar student prefix makes the teacher less certain, and its token supervision less reliable.
🧭 We introduce SCOUT (Student-COnditioned Updates of the Teacher), which adapts the teacher to student prefixes.
(1/10)

[Prithwish Jana @jprithwish]

[7/n]
Huge thanks to my mentors @kanakaris_nikos and @SahikaGenc, my manager @PurakJain, and my amazing collaborators @MononitoGoswami, @hliucola, @RLixinyu, @shrangoh, @Zhehui_Huang, @zhishenhuang, Patrick Blöbaum and @adeoras.

[Prithwish Jana @jprithwish]

[1/n]
🧬 Want a stronger agent without retraining? Give it a better harness.

MILO, our recent work from AWS AI Labs, does this automatically: +12.0% resolution rate on Terminal-Bench 2.1, and new records on 3 open math problems, beating AlphaEvolve, TTT-Discover and EvoX.

[Prithwish Jana @jprithwish]

[6/n]
Work done at AWS AI Labs (@AmazonScience). Check out our paper and project page, with replays of the real runs:
📄 Paper: https://t.co/ZuprpYL5KQ
🔗 Project page: https://t.co/R7a3X6E2Gs