Elamur Research⌕
☰
home / authors / Mehul Damani

Mehul Damani

Mehul Damani has 3 papers in our library, in 2024–2026, most of them on Efficient models & compression; often with Idan Shenfeld, Jacob Andreas, Andi Peng.

3 papers in our library · 2024–2026

OpenAlex · ORCID · Wikidata · h-index 6 · 23 works · 362 citations OpenAlex, 2026-10-14

📄 3 works in our library · 🔗 Cited 140 times in our library · 1 more in our library without a page yet · ⭐ Paper Stars ranking: #7965

The profile is the library's match of the name; two people with one name are not always told apart.

Elsewhere: Google Scholar · Semantic Scholar · dblp · ACL Anthology

 Affiliations

🎓Massachusetts Institute of TechnologyCambridge · United States2023–2026 · 10 paperscurrent🎓National University of SingaporeSingapore · Singapore2021–2022 · 3 papers🔬MIT Computer Science and Artificial Intelligence LaboratoryCambridge · United States2023–2024 · 2 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 1 more institution on single bylines are not listed.

 Fields, topics and concepts

Of its works in our library, with the number of works on each.

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Artificial Intelligence · 3 worksArts and Humanities · 3 worksComputer Science · 3 worksLanguage and Linguistics · 3 works

🧭 Topics 1Research subjects and languages the work is about.Efficient models & compression · 1 work

 Works by year

2020s

20261 paper20242 papers

 Papers

The most cited first (OpenAlex and our library).

  1. Self-Distillation Enables Continual Learning (2026 · 62 citations)
    Foundation models usually stop learning after deployment. Fine-tuning them on new data with supervised fine-tuning (SFT) tends to erase earlier abilities, because SFT trains on a fixed external data distribution, which makes it off-policy. On-policy reinforcement learning forgets less, but it requires a reward…
  2. The Surprising Effectiveness of Test-Time Training for Few-Shot Learning (2024 · 36 citations)
    Language models handle tasks similar to their training data well. However, on structurally novel problems they often fail, even when the prompt contains a few examples. The authors, from MIT, investigate test-time training (TTT). During inference, the model is temporarily fine-tuned through adapters, using data…
  3. Learning How Hard to Think: Input-Adaptive Allocation of LM Computation (2024 · 22 citations)
    Costly decoding strategies such as best-of-k sampling with reranking, or calling a stronger model, improve LM outputs, but they are usually applied uniformly to every query.

All works

4 works of theirs in our library, the most cited first; 3 with a page here, the others link their original until we summarise them.

  1. Self-Distillation Enables Continual LearningarXiv:2601.19897Idan Shenfeld, Mehul Damani, Jonas Hubotter, et al.2026 · 62 citations
  2. The Surprising Effectiveness of Test-Time Training for Few-Shot LearningarXiv:2411.07279Ekin Akyürek, Mehul Damani, Adam Zweiger, et al.2024 · 36 citations
  3. Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty ↗arXiv (Cornell University) · no summary yetMehul Damani et al.2025 · 24 citations
  4. Learning How Hard to Think: Input-Adaptive Allocation of LM ComputationarXiv:2410.04707Mehul Damani, Idan Shenfeld, Andi Peng, et al.2024 · 22 citations

 Co-authors here

8 co-authors with a page, the most papers together first.

 Cited by

Papers of our library cite their works 140 times (of the 149,439 papers whose bibliographies we read); 41 citing papers with a page here →.

Updated

This page in Markdown → https://md.research.elamur.ai/free/en/authors/mehul-damani/