| Efficient Document Ranking with Learnable Late Interactions Z Ji, H Jain, A Veit, SJ Reddi, S Jayasumana, AS Rawat, AK Menon, F Yu, ... arXiv preprint arXiv:2406.17968, 2024 | | 2024 |
| Cascade-Aware Training of Language Models C Wang, S Augenstein, K Rush, W Jitkrittum, H Narasimhan, AS Rawat, ... arXiv preprint arXiv:2406.00060, 2024 | | 2024 |
| Faster Cascades via Speculative Decoding H Narasimhan, W Jitkrittum, AS Rawat, S Kim, N Gupta, AK Menon, ... arXiv preprint arXiv:2405.19261, 2024 | | 2024 |
| Performing classification tasks using post-hoc estimators for expert deferral H Narasimhan, W Jitkrittum, AK Menon, AS Rawat, S Kumar US Patent App. 18/488,951, 2024 | | 2024 |
| Language Model Cascades: Token-level uncertainty and beyond N Gupta, H Narasimhan, W Jitkrittum, AS Rawat, AK Menon, S Kumar arXiv preprint arXiv:2404.10136, 2024 | 4 | 2024 |
| Metric-aware LLM inference M Lukasik, H Narasimhan, AK Menon, F Yu, S Kumar arXiv preprint arXiv:2403.04182, 2024 | 2 | 2024 |
| When does confidence-based cascade deferral suffice? W Jitkrittum, N Gupta, AK Menon, H Narasimhan, A Rawat, S Kumar Advances in Neural Information Processing Systems 36, 2024 | 6 | 2024 |
| ResMem: Learn what you can and memorize the rest Z Yang, M Lukasik, V Nagarajan, Z Li, A Rawat, M Zaheer, AK Menon, ... Advances in Neural Information Processing Systems 36, 2024 | 5 | 2024 |
| On student-teacher deviations in distillation: does it pay to disobey? V Nagarajan, AK Menon, S Bhojanapalli, H Mobahi, S Kumar Advances in Neural Information Processing Systems 36, 5961-6000, 2023 | 4 | 2023 |
| Distillspec: Improving speculative decoding via knowledge distillation Y Zhou, K Lyu, AS Rawat, AK Menon, A Rostamizadeh, S Kumar, JF Kagy, ... arXiv preprint arXiv:2310.08461, 2023 | 29 | 2023 |
| What do larger image classifiers memorise? M Lukasik, V Nagarajan, AS Rawat, AK Menon, S Kumar arXiv preprint arXiv:2310.05337, 2023 | 3 | 2023 |
| Think before you speak: Training language models with pause tokens S Goyal, Z Ji, AS Rawat, AK Menon, S Kumar, V Nagarajan arXiv preprint arXiv:2310.02226, 2023 | 17 | 2023 |
| The importance of feature preprocessing for differentially private linear optimization Z Sun, AT Suresh, AK Menon arXiv preprint arXiv:2307.11106, 2023 | 1 | 2023 |
| Robust distillation for worst-class performance: on the interplay between teacher and student objectives S Wang, H Narasimhan, Y Zhou, S Hooker, M Lukasik, AK Menon Uncertainty in Artificial Intelligence, 2237-2247, 2023 | 7 | 2023 |
| Training machine learning models to be robust against label noise AK Menon, AS Rawat, SJ Reddi, S Kumar US Patent 11,676,033, 2023 | 1 | 2023 |
| Plugin estimators for selective classification with out-of-distribution detection H Narasimhan, AK Menon, W Jitkrittum, S Kumar arXiv preprint arXiv:2301.12386, 2023 | | 2023 |
| Supervision complexity and its role in knowledge distillation H Harutyunyan, AS Rawat, AK Menon, S Kim, S Kumar arXiv preprint arXiv:2301.12245, 2023 | 8 | 2023 |
| Embeddistill: A geometric knowledge distillation for information retrieval S Kim, AS Rawat, M Zaheer, S Jayasumana, V Sadhanala, W Jitkrittum, ... arXiv preprint arXiv:2301.12005, 2023 | 5 | 2023 |
| Accounting for long-tail training data through logit adjustment AK Menon, S Kumar, H Jain, A Veit, AS Rawat, GSJHM Kankanamge US Patent App. 17/375,960, 2023 | | 2023 |
| Learning to reject meets ood detection: Are all abstentions created equal H Narasimhan, AK Menon, W Jitkrittum, S Kumar arXiv preprint arXiv:2301.12386, 2023 | 9 | 2023 |
Sanjiv KumarGoogle Fellow, VP, Google ResearchVerified email at google.com
