Search Results for author:
Found 24 papers, 18 papers with code
CULTURE-GEN: Revealing Global Cultural Perception in Language Models through Natural Language Prompting
As the utilization of large language models (LLMs) has proliferated worldwide, it is crucial for them to have adequate knowledge and fair representation for diverse global cultures.
RewardBench: Evaluating Reward Models for Language Modeling
In this paper, we present RewardBench, a benchmark dataset and code-base for evaluation, to enhance scientific understanding of reward models.
A Roadmap to Pluralistic Alignment
We identify and formalize three possible ways to define and operationalize pluralism in AI systems: 1) Overton pluralistic models that present a spectrum of reasonable responses; 2) Steerably pluralistic models that can steer to reflect certain perspectives; and 3) Distributionally pluralistic models that are well-calibrated to a given population in distribution.
The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning
We analyze the effect of alignment tuning by examining the token distribution shift between base LLMs and their aligned counterpart.
The Generative AI Paradox: "What It Can Create, It May Not Understand"
Specifically, we propose and test the Generative AI Paradox hypothesis: generative models, having been trained directly to reproduce expert-like outputs, acquire generative capabilities that are not contingent upon -- and can therefore exceed -- their ability to understand those same types of outputs.
What Makes it Ok to Set a Fire? Iterative Self-distillation of Contexts and Rationales for Disambiguating Defeasible Social and Moral Situations
From this model we distill a high-quality dataset, \delta-Rules-of-Thumb, of 1. 2M entries of contextualizations and rationales for 115K defeasible moral actions rated highly by human annotators 85. 9% to 99. 8% of the time.
Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinement
The ability to derive underlying principles from a handful of observations and then generalize to novel situations -- known as inductive reasoning -- is central to human intelligence.
Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties
To improve AI systems to better reflect value pluralism, the first-order challenge is to explore the extent to which AI systems can model pluralistic human values, rights, and duties as well as their interaction.
Fine-Grained Human Feedback Gives Better Rewards for Language Model Training
We introduce Fine-Grained RLHF, a framework that enables training and learning from reward functions that are fine-grained in two respects: (1) density, providing a reward after every segment (e. g., a sentence) is generated; and (2) incorporating multiple reward models associated with different feedback types (e. g., factual incorrectness, irrelevance, and information incompleteness).
Faith and Fate: Limits of Transformers on Compositionality
We formulate compositional tasks as computation graphs to systematically quantify the level of complexity, and break down reasoning steps into intermediate sub-procedures.
Inference-Time Policy Adapters (IPA): Tailoring Extreme-Scale LMs without Fine-tuning
While extreme-scale language models have demonstrated exceptional performance on a variety of language tasks, the degree of control over these language models through pure prompting can often be limited.
Evaluating Open-Domain Question Answering in the Era of Large Language Models
The recent success of large language models (LLMs) for QA aggravates lexical matching failures since candidate answers become longer, thereby making matching with the gold answers even more challenging.
Elastic Weight Removal for Faithful and Abstractive Dialogue Generation
We evaluate our method -- using different variants of Flan-T5 as a backbone language model -- on multiple datasets for information-seeking dialogue generation and compare our method with state-of-the-art techniques for faithfulness, such as CTRL, Quark, DExperts, and Noisy Channel reranking.
Self-Refine: Iterative Refinement with Self-Feedback
Motivated by how humans refine their written text, we introduce Self-Refine, an approach for improving initial outputs from LLMs through iterative feedback and refinement.
CHAMPAGNE: Learning Real-world Conversation from Large-Scale Web Videos
To train CHAMPAGNE, we collect and release YTD-18M, a large-scale corpus of 18M video-based dialogues.
FaithDial: A Faithful Benchmark for Information-Seeking Dialogue
The goal of information-seeking dialogue is to respond to seeker queries with natural language utterances that are grounded on knowledge sources.
On the Origin of Hallucinations in Conversational Models: Is it the Datasets or the Models?
Knowledge-grounded conversational models are known to suffer from producing factually invalid statements, a phenomenon commonly called hallucination.
Decomposed Mutual Information Estimation for Contrastive Representation Learning
We propose decomposing the full MI estimation problem into a sum of smaller estimation problems by splitting one of the views into progressively more informed subviews and by applying the chain rule on MI between the decomposed views.
Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark
To this end, we introduce the Benchmark for Evaluation of Grounded INteraction (BEGIN), comprised of 12k dialogue turns generated by neural dialogue systems trained on three knowledge-grounded dialogue corpora.
Neural Path Hunter: Reducing Hallucination in Dialogue Systems via Path Grounding
Dialogue systems powered by large pre-trained language models (LM) exhibit an innate ability to deliver fluent and natural-looking responses.
Decomposing Mutual Information for Representation Learning
In this paper, we transform each view into a set of subviews and then decompose the original MI bound into a sum of bounds involving conditional MI between the subviews.
Evaluating Coherence in Dialogue Systems using Entailment
Evaluating open-domain dialogue systems is difficult due to the diversity of possible correct answers.
Augmenting Neural Response Generation with Context-Aware Topical Attention
Our model is built upon the basic Seq2Seq model by augmenting it with a hierarchical joint attention mechanism that incorporates topical concepts and previous interactions into the response generation.
Automatic Dialogue Generation with Expressed Emotions
Despite myriad efforts in the literature designing neural dialogue generation systems in recent years, very few consider putting restrictions on the response itself.
