Papers by Alona Fyshe
The Emergence of Semantics in Neural Network Representations of Visual Information (N18-2)
Copied to clipboard
| Challenge: | Convolutional neural networks learn about semantics through corpora, but they must be shared . a recent study shows that concepts exist independently of language . |
| Approach: | They employ techniques previously used to detect semantic representations in the human brain to detect representations of CNNs. |
| Outcome: | The proposed techniques could be used to combat adversarial attacks on CNNs, the authors say . |
Characterizing Human and Zero-Shot GPT-3.5 Object-Similarity Judgments (2024.findings-naacl)
Copied to clipboard
| Challenge: | Recent advances in large language models have yielded few-shot, human-comparable performance on a range of tasks, but studies of LLM annotation accuracy and behavior are sparse. |
| Approach: | They characterize OpenAI’s GPT-3.5’s judgment on a behavioral task for implicit object categorization and give similarities and differences between them. |
| Outcome: | The proposed model augments human responses with LLMs for domains where data is sparse or compute resources are low. |
RIFF: Learning to Rephrase Inputs for Few-shot Fine-tuning of Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Pre-trained Language Models (PLMs) can be fine-tuned for downstream text processing tasks. |
| Approach: | They propose to use paraphrases to enrich the input text of a few-shot model with a Maximum-Marginal Likelihood objective to improve performance. |
| Outcome: | The proposed methods improve performance beyond what can be achieved with parameter-efficient fine-tuning alone. |
Weakly-Supervised Questions for Zero-Shot Relation Extraction (2023.eacl-main)
Copied to clipboard
| Challenge: | Zero-Shot Relation Extraction (ZRE) is a task where the training and test sets have no shared relation types. |
| Approach: | They propose to learn a model that can translate relation descriptions into relevant questions, which are then leveraged to generate the correct tail entity. |
| Outcome: | The proposed model outperforms the state-of-the-art on the fewrel and WikiZSL datasets by more than 16 F1 points without using gold question templates. |
Question Generation for Reading Comprehension Assessment by Modeling How and What to Ask (2022.findings-acl)
Copied to clipboard
| Challenge: | Existing Question Generation systems focus on extractive questions and do not control the type of questions. |
| Approach: | They propose a question generation model that generates inferential questions from text . they propose he model can generate questions annotated with story-based reading comprehension skills . |
| Outcome: | The proposed model outperforms baselines on a reading comprehension dataset. |
Offline Preference Optimization via Maximum Marginal Likelihood Estimation (2026.eacl-long)
Copied to clipboard
| Challenge: | Existing approaches to align Large Language Models with human preferences are complex and unstable. |
| Approach: | They propose a new approach that maximizes the marginal log-likelihood of a preferred text output by using the preference pair as samples for approximation. |
| Outcome: | The proposed approach maximizes the marginal log-likelihood of a preferred text output, using the preference pair as samples for approximation, and forgoes the need for both an explicit reward model and entropy maximization. |
Language and Mental Health: Measures of Emotion Dynamics from Text as Linguistic Biosocial Markers (2023.emnlp-main)
Copied to clipboard
| Challenge: | valence variability was significantly lower in the control group compared to ADHD, depression, bipolar disorder, MDD, PTSD, and OCD but not PPD. |
| Approach: | They study the relationship between tweet emotion dynamics and mental health disorders by using a user-disclosed diagnosis. |
| Outcome: | The results show that the measures varied by the user's self-disclosed diagnosis. |
From Language to Language-ish: How Brain-Like is an LSTM’s Representation of Nonsensical Language Stimuli? (2020.findings-emnlp)
Copied to clipboard
| Challenge: | LSTMs are often used to measure event related potentials, but are they able to generalize to new data in a human-like way? |
| Approach: | They asked whether an LSTM model represents a language sample with degraded semantic or syntactic information and whether it resembles the brain's reaction to the stimuli. |
| Outcome: | The results suggest that LSTMs and human brain handle nonsensical data similarly. |