Papers by David Reitter
Investigating Content Planning for Navigating Trade-offs in Knowledge-Grounded Dialogue (2024.eacl-long)
Copied to clipboard
| Challenge: | Knowledge-grounded dialogues require a balance between being specific to what the conversation partner has said and being attributable to an underlying source document. |
| Approach: | They propose a framework that allows to experiment with various plan variables supported by prior work . they show that metric-aware planning mechanisms are better at automatic evaluations but underperform in human judgment compared to metric agnostic mechanisms. |
| Outcome: | The proposed framework supports metric-agnostic and metric aware content planning, but it underperforms in human judgment. |
CONQRR: Conversational Query Rewriting for Retrieval with Reinforcement Learning (2022.emnlp-main)
Copied to clipboard
Zeqiu Wu, Yi Luan, Hannah Rashkin, David Reitter, Hannaneh Hajishirzi, Mari Ostendorf, Gaurav Singh Tomar
| Challenge: | Existing models for conversational question answering require specific retrievers to understand user questions. |
| Approach: | They develop a query rewriting model CONQRR that rewrites a conversational question into a standalone question. |
| Outcome: | The proposed model achieves state-of-the-art on an open-domain conversational question answering dataset and is effective for two different off-the shelf retrievers. |
Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark (2022.tacl-1)
Copied to clipboard
| Challenge: | Knowledge-grounded dialogue systems powered by large language models often generate responses that, while fluent, are not attributable to a relevant source of information. |
| Approach: | They propose to evaluate the validity of 12k dialogue turns generated by neural dialogue systems trained on three knowledge-grounded dialogue corpora and to use them to analyze eight evaluation metrics. |
| Outcome: | The proposed evaluation metrics rely on spurious correlations, do not reliably distinguish attributable abstractive responses from unattributable ones, and perform substantially worse when the knowledge source is longer. |
How do decoding algorithms distribute information in dialogue responses? (2023.findings-eacl)
Copied to clipboard
| Challenge: | Using different decoding algorithms, we find that human dialogue generation is beneficial for adherence to the Uniform Information Density principle. |
| Approach: | They investigate whether decoding algorithms implicitly follow the Uniform Information Density principle by distributing information evenly in utterances. |
| Outcome: | The proposed method encourages non-uniform responses, but under low/high surprisal conditions, resulting in poor quality responses. |
Dungeons and Dragons as a Dialog Challenge for Artificial Intelligence (2022.emnlp-main)
Copied to clipboard
| Challenge: | researchers have posited Dungeons and Dragons as a challenge problem to test systems on various language-related capabilities. |
| Approach: | They frame Dungeons and Dragons specifically as a dialogue system challenge . they train a large language model to generate the next game turn, conditioning it on different information. |
| Outcome: | The proposed game generates the next conversational turn and predicts the state of the game given the dialogue history. |
Like a Baby: Visually Situated Neural Language Acquisition (P19-1)
Copied to clipboard
| Challenge: | A multi-modal neural architecture outperforms its equivalent trained on language alone with a 2% decrease in perplexity . |
| Approach: | They propose to use visual context to train neural language models to perform next-word prediction. |
| Outcome: | The proposed model outperforms its equivalent trained on language with 2% decrease in perplexity even when no visual context is available at test. |
Not that much power: Linguistic alignment is influenced more by low-level linguistic features rather than social power (P18-1)
Copied to clipboard
| Challenge: | linguistic alignment between interlocutors of higher power is attributed to their relative social power, but studies on low-level linguistic features do not account for these factors. |
| Approach: | They characterize the effect of power on alignment with logistic regression models in two datasets and find it vanishes after controlling for low-level features such as utterance length. |
| Outcome: | The proposed model shows that the effect vanishes or is reversed after controlling for low-level features such as utterance length. |
The Timing of Lexical Memory Retrievals in Language Production (N18-1)
Copied to clipboard
| Challenge: | In a large-scale observational study of a spoken corpus, we find that language production at a time point preceding a word is sped up or slowed down depending on activation of that word. |
| Approach: | They propose a cognitive model of fluency in which lexical memory retrievals may explain some of the variability in speech rates. |
| Outcome: | The proposed model predicts that language production is sped up or slowed down depending on activation of a word . |
Increasing Faithfulness in Knowledge-Grounded Dialogue with Controllable Features (2021.acl-long)
Copied to clipboard
| Challenge: | Existing systems that strive to be informative teachers are difficult to build . knowledge grounded dialogue systems are difficult because of limited training objectives . |
| Approach: | They propose to train a generative neural dialogue model that is controlled to stay faithful to evidence . they propose to use additional inputs to generate more objective responses . |
| Outcome: | The proposed model produces responses that are perceived by humans to be objective and faithful to evidence. |
Fusion of Detected Objects in Text for Visual Question Answering (D19-1)
Copied to clipboard
| Challenge: | Recent neural architectures such as Transformer and BERT allow for multimodal context, which may help model the meaning of words in general and also sharpen its understanding of instances of words. |
| Approach: | They propose a neural architecture that combines vision and natural language to advance models of multimodal context. |
| Outcome: | The proposed architecture achieves the highest performance on the Visual Commonsense Reasoning benchmark and the best performance to date on the public leaderboard. |
Surprisal Predicts Code-Switching in Chinese-English Bilingual Text (2020.emnlp-main)
Copied to clipboard
| Challenge: | a new study examines the propensity of bilinguals to switch languages . word surprisal and word entropy are important predictors of code-switching . |
| Approach: | They propose high cognitive effort as a reason for code-switching . they use a computational model of surprisal and word entropy to model code-changing . |
| Outcome: | The proposed model shows that word surprisal, but not entropy, is a significant predictor . sentence length is also a predictor, which has been related to sentence complexity . |