papers

Publications (40)

cs.LG2025

Don't lie to your friends: Learning what you know from collaborative self-play

Jacob Eisenstein, Reza Aghajani, Adam Fisch +5

To be helpful assistants, AI agents must be aware of their own capabilities and limitations. This includes knowing when to answer from parametric knowledge versus using tools, when…

cs.LG2026

Learning Steerable Clarification Policies with Collaborative Self-play

Jonathan Berant, Maximillian Chen, Adam Fisch +4

To handle underspecified or ambiguous queries, AI assistants need a policy for managing their uncertainty to determine (a) when to guess the user intent and answer directly, (b) wh…

cs.CL2024

Conformal Language Modeling

Victor Quach, Adam Fisch, Tal Schuster +4

We propose a novel approach to conformal prediction for generative language models (LMs). Standard conformal prediction produces prediction sets -- in place of single predictions -…

cs.LG2024

Helping or Herding? Reward Model Ensembles Mitigate but do not Eliminate Reward Hacking

Jacob Eisenstein, Chirag Nagpal, Alekh Agarwal +9

Reward models play a key role in aligning language model applications towards human preferences. However, this setup creates an incentive for the language model to exploit errors i…

cs.CL2021

Consistent Accelerated Inference via Confident Adaptive Transformers

Tal Schuster, Adam Fisch, Tommi Jaakkola +1

We develop a novel approach for confidently accelerating inference in the large and expensive multilayer Transformers that are now ubiquitous in natural language processing (NLP).…

cs.CL2019

MRQA 2019 Shared Task: Evaluating Generalization in Reading Comprehension

Adam Fisch, Alon Talmor, Robin Jia +3

We present the results of the Machine Reading for Question Answering (MRQA) 2019 shared task on evaluating the generalization capabilities of reading comprehension systems. In this…

cs.CL2026

Robust LLM Performance Certification via Constrained Maximum Likelihood Estimation

Minghe Shen, Ananth Balashankar, Adam Fisch +2

The ability to rigorously estimate the failure rates of large language models (LLMs) is a prerequisite for their safe deployment. Currently, however, practitioners often face a tra…

cs.CL2018

ParlAI: A Dialog Research Software Platform

Alexander H. Miller, Will Feng, Adam Fisch +5

We introduce ParlAI (pronounced "par-lay"), an open-source software platform for dialog research implemented in Python, available at http://parl.ai. Its goal is to provide a unifie…

cs.LG2024

Stratified Prediction-Powered Inference for Hybrid Language Model Evaluation

Adam Fisch, Joshua Maynez, R. Alex Hofer +3

Prediction-powered inference (PPI) is a method that improves statistical estimates based on limited human-labeled data. PPI achieves this by combining small amounts of human-labele…

cs.LG2024

Bayesian Prediction-Powered Inference

R. Alex Hofer, Joshua Maynez, Bhuwan Dhingra +3

Prediction-powered inference (PPI) is a method that improves statistical estimates based on limited human-labeled data. Specifically, PPI methods provide tighter confidence interva…

cs.LG2025

Plantain: Plan-Answer Interleaved Reasoning

Anthony Liang, Jonathan Berant, Adam Fisch +3

Reasoning models often spend a significant amount of time thinking before they generate a visible response. In the meantime, they do not give the user any hints as to whether their…

cs.LG2021

Few-shot Conformal Prediction with Auxiliary Tasks

Adam Fisch, Tal Schuster, Tommi Jaakkola +1

We develop a novel approach to conformal prediction when the target task has limited data available for training. Conformal prediction identifies a small set of promising output ca…

cs.AI2026

Rich Insights from Cheap Signals: Efficient Evaluations via Tensor Factorization

Felipe Maia Polo, Aida Nematzadeh, Virginia Aglietti +2

Moving beyond evaluations that collapse performance across heterogeneous prompts toward fine-grained evaluation at the prompt level, or within relatively homogeneous subsets, is ne…

cs.CL2019

Working Hard or Hardly Working: Challenges of Integrating Typology into Neural Dependency Parsers

Adam Fisch, Jiang Guo, Regina Barzilay

This paper explores the task of leveraging typology in the context of cross-lingual dependency parsing. While this linguistic information has shown great promise in pre-neural pars…

cs.LG2023

Risk-Controlling Model Selection via Guided Bayesian Optimization

Bracha Laufer-Goldshtein, Adam Fisch, Regina Barzilay +1

Adjustable hyperparameters of machine learning models typically impact various key trade-offs such as accuracy, fairness, robustness, or inference cost. Our goal in this paper is t…

cs.CL2026

MT-PingEval: Evaluating Multi-Turn Collaboration with Private Information Games

Jacob Eisenstein, Fantine Huot, Adam Fisch +2

We present a scalable methodology for evaluating language models in multi-turn interactions, using a suite of collaborative games that require effective communication about private…

cs.CL2017

StarSpace: Embed All The Things!

Ledell Wu, Adam Fisch, Sumit Chopra +3

We present StarSpace, a general-purpose neural embedding model that can solve a wide variety of problems: labeling tasks such as text classification, ranking tasks such as informat…

stat.ME2025

Conformal Risk Control

Anastasios N. Angelopoulos, Stephen Bates, Adam Fisch +2

We extend conformal prediction to control the expected value of any monotone loss function. The algorithm generalizes split conformal prediction together with its coverage guarante…

cs.LG2024

Mitigating LLM Hallucinations via Conformal Abstention

Yasin Abbasi Yadkori, Ilja Kuzborskij, David Stutz +9

We develop a principled procedure for determining when a large language model (LLM) should abstain from responding (e.g., by saying "I don't know") in a general domain, instead of…

cs.LG2021

Efficient Conformal Prediction via Cascaded Inference with Expanded Admission

Adam Fisch, Tal Schuster, Tommi Jaakkola +1

In this paper, we present a novel approach for conformal prediction (CP), in which we aim to identify a set of promising prediction candidates -- in place of a single prediction. T…

cs.LG2025

Robust Preference Optimization through Reward Model Distillation

Adam Fisch, Jacob Eisenstein, Vicky Zayats +5

Language model (LM) post-training (or alignment) involves maximizing a reward function that is derived from preference annotations. Direct Preference Optimization (DPO) is a popula…

cs.LG2022

Efficiently Controlling Multiple Risks with Pareto Testing

Bracha Laufer-Goldshtein, Adam Fisch, Regina Barzilay +1

Machine learning applications frequently come with multiple diverse objectives and constraints that can change over time. Accordingly, trained models can be tuned with sets of hype…

cs.LG2022

Conformal Prediction Sets with Limited False Positives

Adam Fisch, Tal Schuster, Tommi Jaakkola +1

We develop a new approach to multi-label conformal prediction in which we aim to output a precise set of promising prediction candidates with a bounded number of incorrect answers.…

cs.LG2024

Calibrated Selective Classification

Adam Fisch, Tommi Jaakkola, Regina Barzilay

Selective classification allows models to abstain from making predictions (e.g., say "I don't know") when in doubt in order to obtain better effective accuracy. While typical selec…

cs.CL2023

Towards Robust and Efficient Continual Language Learning

Adam Fisch, Amal Rannen-Triki, Razvan Pascanu +4

As the application space of language models continues to evolve, a natural question to ask is how we can quickly adapt models to new tasks. We approach this classic question from a…

cs.CL2016

Key-Value Memory Networks for Directly Reading Documents

Alexander Miller, Adam Fisch, Jesse Dodge +3

Directly reading documents and being able to answer questions from them is an unsolved challenge. To avoid its inherent difficulty, question answering (QA) has been directed toward…

cs.CL2024

Block Transformer: Global-to-Local Language Modeling for Fast Inference

Namgyu Ho, Sangmin Bae, Taehyeon Kim +6

We introduce the Block Transformer which adopts hierarchical global-to-local modeling to autoregressive transformers to mitigate the inference bottlenecks associated with self-atte…

cs.LG2026

CollabEval: Statistically Efficient Collaborative Model Evaluation via Matrix Completion

Adam Fisch, Daniel Deutsch, Joshua Maynez +5

Evaluating generative AI models is a routine, but resource-intensive, process that is conducted over and over again during the course of model development. In this work, we propose…

cs.CL2017

Reading Wikipedia to Answer Open-Domain Questions

Danqi Chen, Adam Fisch, Jason Weston +1

This paper proposes to tackle open- domain question answering using Wikipedia as the unique knowledge source: the answer to any factoid question is a text span in a Wikipedia artic…

cs.LG2025

Cost-Optimal Active AI Model Evaluation

Anastasios N. Angelopoulos, Jacob Eisenstein, Jonathan Berant +2

The development lifecycle of generative AI systems requires continual evaluation, data acquisition, and annotation, which is costly in both resources and time. In practice, rapid i…

math.ST2026

Multiple-Prediction-Powered Inference

Charlie Cowen-Breen, Alekh Agarwal, Stephen Bates +4

Statistical estimation often involves tradeoffs between expensive, high-quality measurements and a variety of lower-quality proxies. We introduce Multiple-Prediction-Powered Infere…

cs.CL2023

Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models

Aarohi Srivastava, Abhinav Rastogi, Abhishek Rao +448

Language models demonstrate both quantitative improvement and new qualitative capabilities with increasing scale. Despite their potentially transformative impact, these new capabil…

cs.LG2024

Rewarding Progress: Scaling Automated Process Verifiers for LLM Reasoning

Amrith Setlur, Chirag Nagpal, Adam Fisch +6

A promising approach for improving reasoning in large language models is to use process reward models (PRMs). PRMs provide feedback at each step of a multi-step reasoning trace, po…

cs.CL2025

Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA

Sangmin Bae, Adam Fisch, Hrayr Harutyunyan +3

Large language models (LLMs) are expensive to deploy. Parameter sharing offers a possible path towards reducing their size and cost, but its effectiveness in modern LLMs remains fa…

cs.CL2025

Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation

Sangmin Bae, Yujin Kim, Reza Bayat +8

Scaling language models unlocks impressive capabilities, but the accompanying computational and memory demands make both training and deployment expensive. Existing efficiency effo…

cs.CL2021

Making Pre-trained Language Models Better Few-shot Learners

Tianyu Gao, Adam Fisch, Danqi Chen

The recent GPT-3 model (Brown et al., 2020) achieves remarkable few-shot performance solely by leveraging a natural-language prompt and a few task demonstrations as input context.…

cs.CL2020

CapWAP: Captioning with a Purpose

Adam Fisch, Kenton Lee, Ming-Wei Chang +2

The traditional image captioning task uses generic reference captions to provide textual information about images. Different user populations, however, will care about different vi…

cs.CL2022

Confident Adaptive Language Modeling

Tal Schuster, Adam Fisch, Jai Gupta +5

Recent advances in Transformer-based large language models (LLMs) have led to significant performance improvements across many tasks. These gains come with a drastic increase in th…

cs.LG2025

Mitigating Preference Hacking in Policy Optimization with Pessimism

Dhawal Gupta, Adam Fisch, Christoph Dann +1

This work tackles the problem of overoptimization in reinforcement learning from human feedback (RLHF), a prevalent technique for aligning models with human preferences. RLHF relie…

cs.CL2021

Get Your Vitamin C! Robust Fact Verification with Contrastive Evidence

Tal Schuster, Adam Fisch, Regina Barzilay

Typical fact verification models use retrieved written evidence to verify claims. Evidence sources, however, often change over time as more information is gathered and revised. In…