Papers by Yinglin Wang
Enhancing the Context Representation in Similarity-based Word Sense Disambiguation (2021.emnlp-main)
Copied to clipboard
| Challenge: | Existing similarity-based systems focus on learning sense embeddings using only the sentence where the word appears, neglecting its global context. |
| Approach: | They propose a contextoriented embedding technique that takes better advantage of both word-level and sense-level global context of an ambiguous word for disambiguation. |
| Outcome: | The proposed method improves on all-words WSD benchmarks in knowledge-based category by large margins. |
Doing Good or Doing Right? Exploring the Weakness of Commonsense Causal Reasoning Models (2021.acl-short)
Copied to clipboard
| Challenge: | Pretrained language models (PLMs) achieve surprising performance on the Choice of Plausible Alternatives (COPA) task. |
| Approach: | They propose to add a regularization loss to the existing COPA models to mitigate the problem of semantic similarity bias by adding a normalization loss. |
| Outcome: | The proposed model improves generalization ability and performs better on a challenging dataset, BCOPA-CE, which has unbiased token distribution and is more difficult for models to distinguish cause and effect. |
A Synset Relation-enhanced Framework with a Try-again Mechanism for Word Sense Disambiguation (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing sense embeddings fail to embed sense knowledge in semantic networks. |
| Approach: | They propose a Synset Relation-Enhanced Framework that leverages sense relations for sense embedding enhancement and a try-again mechanism that implements WSD again. |
| Outcome: | The proposed system outperforms knowledge-based systems with 20% SemCor data on all-words and lexical datasets. |
V-RoLoRA: RLVR-Driven MoE Routing for Steerable Pluralistic Alignment (2026.findings-acl)
Copied to clipboard
| Challenge: | Current methods for steering large language models rely on prompt engineering or reasoning-time guidance. |
| Approach: | They propose a value-controllable pluralistic alignment framework enhanced with conditioned gating that dynamically directs the flow among multiple experts based on an input value or moral vector. |
| Outcome: | The proposed method outperforms prompt-based steering and multi-task PEFT benchmarks on two 8-billion-parameter backbones. |
Word Sense Disambiguation: Towards Interactive Context Exploitation from Both Word and Sense Perspectives (2021.acl-long)
Copied to clipboard
| Challenge: | Recent Word Sense Disambiguation systems have approached the upper bound of the task on standard evaluation benchmarks. |
| Approach: | They propose to convert the nearly isolated decisions into interrelated ones by exposing senses in context when learning sense embeddings in a similarity-based Sense Aware Context Exploitation architecture. |
| Outcome: | The proposed approach surpasses state-of-the-art on English and multilingual datasets by large margins. |