Author

Tirthankar Ghosal

h-index 13 View on Semantic Scholar

We have 2 of 72 papers

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

Rethinking Domain Specialization for Open-Ended Scientific Reasoning in Astronomy Language Models

This work compares open-weight and API-served general-purpose, multimodal, and astronomy-specialized models using judge-based correctness and complementary reference metrics and highlights the role of domain-specific evaluation in determining which models, capabilities, and evaluation criteria are appropriate for scien...

Vanessa Lama, Sanjay Das, Emily J. Herron et al. · 0 citations

Preprint Aug 2026

Do LLMs Know a Good Hypothesis When They See One? Logit-Based Energy Scoring Outperforms Prompted LLM-as-Judge for Scientific Hypothesis Ranking

A logit-based energy scoring method is proposed that evaluates hypotheses using a language model's intrinsic confidence rather than comparative judgment, and intrinsic model confidence shows potential for scientific hypothesis evaluation.

Swati Rajwal, Sanjay Das, Tirthankar Ghosal · 1 citation