Search Results for author: Ximing Lu

Found 35 papers, 24 papers with code

Localized Symbolic Knowledge Distillation for Visual Commonsense Models

no code implementations NeurIPS 2023 Jae Sung Park, Jack Hessel, Khyathi Raghavi Chandu, Paul Pu Liang, Ximing Lu, Peter West, Youngjae Yu, Qiuyuan Huang, Jianfeng Gao, Ali Farhadi, Yejin Choi

Empirical results and human evaluations in a zero-shot setup demonstrate that our distillation method results in more precise VL models of reasoning compared to a baseline of passing a generated referring expression to an LLM.

Instruction Following Knowledge Distillation +3

The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning

no code implementations4 Dec 2023 Bill Yuchen Lin, Abhilasha Ravichander, Ximing Lu, Nouha Dziri, Melanie Sclar, Khyathi Chandu, Chandra Bhagavatula, Yejin Choi

We analyze the effect of alignment tuning by examining the token distribution shift between base LLMs and their aligned counterpart.

STEER: Unified Style Transfer with Expert Reinforcement

no code implementations13 Nov 2023 Skyler Hallinan, Faeze Brahman, Ximing Lu, JaeHun Jung, Sean Welleck, Yejin Choi

We propose STEER: Unified Style Transfer with Expert Reinforcement, a unified frame-work developed to overcome the challenge of limited parallel data for style transfer.

Style Transfer Text Style Transfer

Tailoring Self-Rationalizers with Multi-Reward Distillation

1 code implementation6 Nov 2023 Sahana Ramnath, Brihi Joshi, Skyler Hallinan, Ximing Lu, Liunian Harold Li, Aaron Chan, Jack Hessel, Yejin Choi, Xiang Ren

Results on five difficult question-answering datasets StrategyQA, QuaRel, OpenBookQA, NumerSense and QASC show that not only does MaRio improve task accuracy, but it also improves the self-rationalization quality of small LMs across the aforementioned axes better than a supervised fine-tuning (SFT) baseline.

Question Answering StrategyQA

The Generative AI Paradox: "What It Can Create, It May Not Understand"

no code implementations31 Oct 2023 Peter West, Ximing Lu, Nouha Dziri, Faeze Brahman, Linjie Li, Jena D. Hwang, Liwei Jiang, Jillian Fisher, Abhilasha Ravichander, Khyathi Chandu, Benjamin Newman, Pang Wei Koh, Allyson Ettinger, Yejin Choi

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.

Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinement

1 code implementation12 Oct 2023 Linlu Qiu, Liwei Jiang, Ximing Lu, Melanie Sclar, Valentina Pyatkin, Chandra Bhagavatula, Bailin Wang, Yoon Kim, Yejin Choi, Nouha Dziri, Xiang Ren

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

1 code implementation2 Sep 2023 Taylor Sorensen, Liwei Jiang, Jena Hwang, Sydney Levine, Valentina Pyatkin, Peter West, Nouha Dziri, Ximing Lu, Kavel Rao, Chandra Bhagavatula, Maarten Sap, John Tasioulas, Yejin Choi

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.

Decision Making

Faith and Fate: Limits of Transformers on Compositionality

1 code implementation NeurIPS 2023 Nouha Dziri, Ximing Lu, Melanie Sclar, Xiang Lorraine Li, Liwei Jiang, Bill Yuchen Lin, Peter West, Chandra Bhagavatula, Ronan Le Bras, Jena D. Hwang, Soumya Sanyal, Sean Welleck, Xiang Ren, Allyson Ettinger, Zaid Harchaoui, Yejin Choi

We formulate compositional tasks as computation graphs to systematically quantify the level of complexity, and break down reasoning steps into intermediate sub-procedures.

Impossible Distillation: from Low-Quality Model to High-Quality Dataset & Model for Summarization and Paraphrasing

no code implementations26 May 2023 JaeHun Jung, Peter West, Liwei Jiang, Faeze Brahman, Ximing Lu, Jillian Fisher, Taylor Sorensen, Yejin Choi

It is commonly perceived that the strongest language models (LMs) rely on a combination of massive scale, instruction data, and human feedback to perform specialized tasks -- e. g. summarization and paraphrasing, without supervision.

Improving Language Models with Advantage-based Offline Policy Gradients

1 code implementation24 May 2023 Ashutosh Baheti, Ximing Lu, Faeze Brahman, Ronan Le Bras, Maarten Sap, Mark Riedl

We introduce Advantage-Leftover Lunch RL (A-LoL), a new class of offline policy gradient algorithms that enable RL training on any pre-existing data.

Language Modelling Offline RL +1

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

1 code implementation CVPR 2023 Youngjae Yu, Jiwan Chung, Heeseung Yun, Jack Hessel, Jae Sung Park, Ximing Lu, Rowan Zellers, Prithviraj Ammanabrolu, Ronan Le Bras, Gunhee Kim, Yejin Choi

Language models are capable of commonsense reasoning: while domain-specific models can learn from explicit knowledge (e. g. commonsense graphs [6], ethical norms [25]), and larger models like GPT-3 manifest broad commonsense reasoning capacity.

Language Modelling reinforcement-learning +2

I2D2: Inductive Knowledge Distillation with NeuroLogic and Self-Imitation

no code implementations19 Dec 2022 Chandra Bhagavatula, Jena D. Hwang, Doug Downey, Ronan Le Bras, Ximing Lu, Lianhui Qin, Keisuke Sakaguchi, Swabha Swayamdipta, Peter West, Yejin Choi

Here, we investigate an alternative that a priori seems impossible: can smaller language models (e. g., GPT-2) win over models that are orders of magnitude larger and better (e. g., GPT-3), if powered with novel commonsense distillation algorithms?

Imitation Learning Knowledge Distillation

Generating Sequences by Learning to Self-Correct

no code implementations31 Oct 2022 Sean Welleck, Ximing Lu, Peter West, Faeze Brahman, Tianxiao Shen, Daniel Khashabi, Yejin Choi

Sequence generation applications require satisfying semantic constraints, such as ensuring that programs are correct, using certain keywords, or avoiding undesirable content.

Language Modelling Program Synthesis

Rainier: Reinforced Knowledge Introspector for Commonsense Question Answering

1 code implementation6 Oct 2022 Jiacheng Liu, Skyler Hallinan, Ximing Lu, Pengfei He, Sean Welleck, Hannaneh Hajishirzi, Yejin Choi

Our work is the first to report that knowledge generated by models that are orders of magnitude smaller than GPT-3, even without direct supervision on the knowledge itself, can exceed the quality of commonsense knowledge elicited from GPT-3.

Question Answering Reinforcement Learning (RL)

NaturalProver: Grounded Mathematical Proof Generation with Language Models

1 code implementation25 May 2022 Sean Welleck, Jiacheng Liu, Ximing Lu, Hannaneh Hajishirzi, Yejin Choi

Theorem proving in natural mathematical language - the mixture of symbolic and natural language used by humans - plays a central role in mathematical advances and education, and tests aspects of reasoning that are core to intelligence.

Automated Theorem Proving Language Modelling

ProsocialDialog: A Prosocial Backbone for Conversational Agents

1 code implementation25 May 2022 Hyunwoo Kim, Youngjae Yu, Liwei Jiang, Ximing Lu, Daniel Khashabi, Gunhee Kim, Yejin Choi, Maarten Sap

With this dataset, we introduce a dialogue safety detection module, Canary, capable of generating RoTs given conversational context, and a socially-informed dialogue agent, Prost.

Dialogue Generation Dialogue Safety Prediction +2

Twist Decoding: Diverse Generators Guide Each Other

1 code implementation19 May 2022 Jungo Kasai, Keisuke Sakaguchi, Ronan Le Bras, Hao Peng, Ximing Lu, Dragomir Radev, Yejin Choi, Noah A. Smith

Our extensive evaluations on machine translation and scientific paper summarization demonstrate that Twist decoding substantially outperforms each model decoded in isolation over various scenarios, including cases where domain-specific and general-purpose models are both available.

Machine Translation Text Generation +1

Connecting the Dots between Audio and Text without Parallel Data through Visual Knowledge Transfer

1 code implementation NAACL 2022 Yanpeng Zhao, Jack Hessel, Youngjae Yu, Ximing Lu, Rowan Zellers, Yejin Choi

In a difficult zero-shot setting with no paired audio-text data, our model demonstrates state-of-the-art zero-shot performance on the ESC50 and US8K audio classification tasks, and even surpasses the supervised state of the art for Clotho caption retrieval (with audio queries) by 2. 2\% R@1.

Audio Classification Audio Tagging +3

NeuroLogic A*esque Decoding: Constrained Text Generation with Lookahead Heuristics

1 code implementation NAACL 2022 Ximing Lu, Sean Welleck, Peter West, Liwei Jiang, Jungo Kasai, Daniel Khashabi, Ronan Le Bras, Lianhui Qin, Youngjae Yu, Rowan Zellers, Noah A. Smith, Yejin Choi

To enable constrained generation, we build on NeuroLogic decoding (Lu et al., 2021), combining its flexibility in incorporating logical constraints with A*esque estimates of future constraint satisfaction.

Machine Translation Table-to-Text Generation

Generated Knowledge Prompting for Commonsense Reasoning

1 code implementation ACL 2022 Jiacheng Liu, Alisa Liu, Ximing Lu, Sean Welleck, Peter West, Ronan Le Bras, Yejin Choi, Hannaneh Hajishirzi

It remains an open question whether incorporating external knowledge benefits commonsense reasoning while maintaining the flexibility of pretrained sequence models.

Language Modelling Open-Ended Question Answering

MERLOT: Multimodal Neural Script Knowledge Models

1 code implementation NeurIPS 2021 Rowan Zellers, Ximing Lu, Jack Hessel, Youngjae Yu, Jae Sung Park, Jize Cao, Ali Farhadi, Yejin Choi

As humans, we understand events in the visual world contextually, performing multimodal reasoning across time to make inferences about the past, present, and future.

Visual Commonsense Reasoning

On-the-Fly Attention Modulation for Neural Generation

no code implementations Findings (ACL) 2021 Yue Dong, Chandra Bhagavatula, Ximing Lu, Jena D. Hwang, Antoine Bosselut, Jackie Chi Kit Cheung, Yejin Choi

Despite considerable advancements with deep neural language models (LMs), neural text generation still suffers from degeneration: the generated text is repetitive, generic, self-contradictory, and often lacks commonsense.

Language Modelling Text Generation

Analyzing Commonsense Emergence in Few-shot Knowledge Models

1 code implementation AKBC 2021 Jeff Da, Ronan Le Bras, Ximing Lu, Yejin Choi, Antoine Bosselut

Our results show that commonsense knowledge models can rapidly adapt from limited examples, indicating that KG fine-tuning serves to learn an interface to encoded knowledge learned during pretraining.

NeuroLogic Decoding: (Un)supervised Neural Text Generation with Predicate Logic Constraints

no code implementations NAACL 2021 Ximing Lu, Peter West, Rowan Zellers, Ronan Le Bras, Chandra Bhagavatula, Yejin Choi

While the dominant recipe for conditional text generation has been large-scale pretrained language models that are finetuned on the task-specific training data, such models do not learn to follow the underlying constraints reliably, even when supervised with large amounts of task-specific examples.

Conditional Text Generation

Reflective Decoding: Beyond Unidirectional Generation with Off-the-Shelf Language Models

no code implementations ACL 2021 Peter West, Ximing Lu, Ari Holtzman, Chandra Bhagavatula, Jena Hwang, Yejin Choi

In this paper, we present Reflective Decoding, a novel unsupervised algorithm that allows for direct application of unidirectional LMs to non-sequential tasks.

Conditional Text Generation Text Infilling

HATNet: An End-to-End Holistic Attention Network for Diagnosis of Breast Biopsy Images

1 code implementation25 Jul 2020 Sachin Mehta, Ximing Lu, Donald Weaver, Joann G. Elmore, Hannaneh Hajishirzi, Linda Shapiro

HATNet extends the bag-of-words approach and uses self-attention to encode global information, allowing it to learn representations from clinically relevant tissue structures without any explicit supervision.

Histopathological Image Classification Image Classification

Cannot find the paper you are looking for? You can Submit a new open access paper.