Christopher Potts

papers

3,047

total citations

papers (14)

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

ICLR 2025arXiv

2,226

citations

Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking

NEURIPS 2021arXiv

citations

CEBaB: Estimating the Causal Effects of Real-World Concepts on NLP Model Behavior

NEURIPS 2022arXiv

citations

Decrypting Cryptic Crosswords: Semantically Complex Wordplay Puzzles as a Target for NLP

NEURIPS 2021arXiv

citations

MrT5: Dynamic Token Merging for Efficient Byte-level Language Models

ICLR 2025arXiv

citations

Base Models Beat Aligned Models at Randomness and Creativity

COLM 2025arXiv

citations

Bayesian scaling laws for in-context learning

COLM 2025arXiv

citations

GIO: Gradient Information Optimization for Training Dataset Selection

ICLR 2024arXiv

citations

ContextRef: Evaluating Referenceless Metrics for Image Description Generation

ICLR 2024arXiv

citations

Blackbox Model Provenance via Palimpsestic Membership Inference

NEURIPS 2025arXiv

citations

Christopher Potts

papers (14)

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

Causal Abstractions of Neural Networks

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

Interpretability at Scale: Identifying Causal Mechanisms in Alpaca

Baleen: Robust Multi-Hop Reasoning at Scale via Condensed Retrieval

Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking

CEBaB: Estimating the Causal Effects of Real-World Concepts on NLP Model Behavior

Decrypting Cryptic Crosswords: Semantically Complex Wordplay Puzzles as a Target for NLP

MrT5: Dynamic Token Merging for Efficient Byte-level Language Models

Base Models Beat Aligned Models at Randomness and Creativity

Bayesian scaling laws for in-context learning

GIO: Gradient Information Optimization for Training Dataset Selection

ContextRef: Evaluating Referenceless Metrics for Image Description Generation

Blackbox Model Provenance via Palimpsestic Membership Inference

papers (14)

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

Causal Abstractions of Neural Networks

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

Interpretability at Scale: Identifying Causal Mechanisms in Alpaca

Baleen: Robust Multi-Hop Reasoning at Scale via Condensed Retrieval

Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarking

CEBaB: Estimating the Causal Effects of Real-World Concepts on NLP Model Behavior

Decrypting Cryptic Crosswords: Semantically Complex Wordplay Puzzles as a Target for NLP

MrT5: Dynamic Token Merging for Efficient Byte-level Language Models

Base Models Beat Aligned Models at Randomness and Creativity

Bayesian scaling laws for in-context learning

GIO: Gradient Information Optimization for Training Dataset Selection

ContextRef: Evaluating Referenceless Metrics for Image Description Generation

Blackbox Model Provenance via Palimpsestic Membership Inference