Author

Nafiul Haque

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Sep 2026

Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning?

Goal-conditioned reinforcement learning (GCRL) relies heavily on how target goals are represented to the policy. While recent methods encode goals via temporal distance, occupancy, or controllability, it remains unclear how much downstream performance actually depends on representation quality. We study this in offline...

Syed Nazmus Sakib, Abdul Monaf Chowdhury, Nafiul Haque et al. · 0 citations

#artificial intelligence Preprint Sep 2026

Unlearnable, or Unmeasured? On the Reliability of Difficulty Labels in RLVR

Reinforcement learning with verifiable rewards (RLVR) has become an important approach for improving reasoning during post-training. Recent work suggests that some difficult prompts remain resistant to learning even when they occasionally produce correct solutions. We revisit this unlearnability phenomenon and find tha...

Chandak Chakma, Syed Nazmus Sakib, Nafiul Haque et al. · 0 citations