🧑‍🔬 Anirudh Goyal

🏛️ Affiliations

🎓Princeton UniversityPrinceton · United States2023–2026 · 4 paperscurrent🎓University of Maryland, College ParkCollege Park · United States2026 · 1 papercurrent🎓Université de MontréalMontreal · Canada2016–2024 · 44 papers🔬Mila - Quebec Artificial Intelligence InstituteMontreal · Canada2016–2024 · 18 papers🏢Google DeepMindLondon · United Kingdom2022–2024 · 8 papers🏢MetaLondon · United Kingdom2025 · 2 papers🎓University of WashingtonSeattle · United States2025 · 2 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 5 more institutions on single bylines are not listed.

🧭 Fields, topics and concepts

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

📅 Works by year

2020s

2010s

📄 Papers

The most cited first (OpenAlex and our library).

1. An Actor-Critic Algorithm for Sequence Prediction (2016 · 335 citations)
   RNN sequence generators are usually trained with teacher forcing. They learn to predict each token from the gold prefix, but at test time they must continue from their own, possibly wrong, outputs. They also optimize log-likelihood rather than the evaluation metric, such as BLEU. The authors adapt…
2. Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem Solving (2024 · 45 citations)
   The authors ask whether large language models have something like metacognitive knowledge, meaning an awareness of which skills a task calls for, and whether that awareness can be used to improve their performance.
3. AI-Assisted Generation of Difficult Math Questions (2024 · 23 citations)
   LLM training now treats mathematical reasoning as a core capability, but public sources of hard problems are largely used up.
4. SWE-Search: Enhancing Software Agents with Monte Carlo Tree Search and Iterative Refinement (2024 · 22 citations)
   Most LLM-based coding agents move through a task in a single forward pass. They cannot easily abandon a poor line of attack and try a different one. SWE-Search tackles this for repository-level bug fixing by combining Monte Carlo Tree Search (MCTS) with LLM self-evaluation. It builds on the moatless-tools framework…
5. Auto-Discovery-Bench: Diagnosing Structured State Tracking in Oracle-Guided Discovery (2025 · 6 citations)
   Agents meant to help with scientific discovery must keep a structured picture of a hidden system accurate across many rounds of experiments and feedback.

All works

12 works of theirs in our library, the most cited first; 5 with a page here, the others link their original until we summarise them.

1. An Actor-Critic Algorithm for Sequence PredictionarXiv:1607.07086Dzmitry Bahdanau, Philemon Brakel, Kelvin Xu, et al.2016 · 335 citations
2. Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem SolvingarXiv:2405.12205Aniket Didolkar, Anirudh Goyal, Nan Rosemary Ke, et al.2024 · 45 citations
3. AI-Assisted Generation of Difficult Math QuestionsarXiv:2407.21009Vedant Shah, Dingli Yu, Kaifeng Lyu, et al.2024 · 23 citations
4. SWE-Search: Enhancing Software Agents with Monte Carlo Tree Search and Iterative RefinementarXiv:2410.20285Antonis Antoniades, Albert Orwall, Kexun Zhang, et al.2024 · 22 citations
5. Learning Dynamics Model in Reinforcement Learning by Incorporating the Long Term Future ↗arXiv (Cornell University) · no summary yetNan Rosemary Ke et al.2019 · 15 citations
6. Modeling the Long Term Future in Model-Based Reinforcement Learning ↗International Conference on Learning Representations · no summary yetNan Rosemary Ke et al.2018 · 14 citations
7. Top-k Training of GANs: Improving GAN Performance by Throwing Away Bad Samples ↗arXiv (Cornell University) · no summary yetSamarth Sinha et al.2020 · 13 citations
8. Rethinking Thinking Tokens: LLMs as Improvement Operators ↗arXiv (Cornell University) · no summary yetLovish Madaan et al.2025 · 8 citations
9. Auto-Discovery-Bench: Diagnosing Structured State Tracking in Oracle-Guided DiscoveryarXiv:2502.15224Tingting Chen, Beibei Lin, Srinivas Anumasa, et al.2025 · 6 citations
10. Scaling Test-Time Compute for Agentic Coding ↗arXiv (Cornell University) · no summary yetJoongwon Kim et al.2026 · 3 citations
11. Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning ↗arXiv (Cornell University) · no summary yetParas Dahal et al.2026
12. GitSwarm: Decentralized Compounding Inference ↗arXiv (Cornell University) · no summary yetVedant Shah et al.2026

🤝 Co-authors here

12 of 21 co-authors with a page, the most papers together first.

• Yoshua Bengio ×3
• Michael C. Mozer ×2
• Nan Rosemary Ke ×2
• Sanjeev Arora ×2
• Vedant Shah ×2
• Aaron Courville
• Aniket Didolkar
• Danilo Jimenez Rezende
• Dianbo Liu
• Dzmitry Bahdanau
• Jiatong Yu
• Joelle Pineau

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

• International Conference on Learning Representations

🔗 Cited by

Papers of our library cite their works 10,174 times (of the 149,439 papers whose bibliographies we read); 2,309 citing papers with a page here →.

Updated 3 November 2026

↑ back to top

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