Yejin Choi

38

Papers

660

Total Citations

Papers (38)

Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory

WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild

ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models

Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties

One-Minute Video Generation with Test-Time Training

Trust or Escalate: LLM Judges with Provable Guarantees for Human Agreement

AI as Humanity’s Salieri: Quantifying Linguistic Creativity of Language Models via Systematic Attribution of Machine Text against Web Text

Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence

SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior

PlaSma: Procedural Knowledge Models for Language-based Planning and Re-Planning

Broken Tokens? Your Language Model can Secretly Handle Non-Canonical Tokenizations

CHAMPAGNE: Learning Real-world Conversation from Large-Scale Web Videos

VisualCOMET: Reasoning about the Dynamic Context of a Still Image

Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks

The Abduction of Sherlock Holmes: A Dataset for Visual Abductive Reasoning

Synthetic Visual Genome

Bias in Gender Bias Benchmarks: How Spurious Features Distort Evaluation

Structured Chemistry Reasoning with Large Language Models

Position: A Roadmap to Pluralistic Alignment

Neural Motifs: Scene Graph Parsing With Global Context

From Recognition to Cognition: Visual Commonsense Reasoning

Tactical Rewind: Self-Correction via Backtracking in Vision-And-Language Navigation

VinVL: Revisiting Visual Representations in Vision-Language Models

MERLOT Reserve: Neural Script Knowledge Through Vision and Language and Sound

Fusing Pre-Trained Language Models With Multimodal Prompts Through Reinforcement Learning

Segment-Phrase Table for Semantic Segmentation, Visual Entailment and Paraphrasing

Defending Against Neural Fake News

MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers

Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals

MERLOT: Multimodal Neural Script Knowledge Models

NaturalProver: Grounded Mathematical Proof Generation with Language Models

COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics

NeurIPS 2022arXiv

QUARK: Controllable Text Generation with Reinforced Unlearning

Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text

Localized Symbolic Knowledge Distillation for Visual Commonsense Models

SwiftSage: A Generative Agent with Fast and Slow Thinking for Complex Interactive Tasks

RealTime QA: What's the Answer Right Now?

Faith and Fate: Limits of Transformers on Compositionality