Papers by Jingcheng Niu

8 papers
Rationally Reappraising ATIS-based Dialogue Systems (P19-1)

Copied to clipboard

Challenge: Recent state-of-the-art neural models have obtained F1-scores near 98% on the task of slot filling.
Approach: They propose to fix annotation errors in ATIS and propose a rule-based grammar for slot filling that achieves a 95.82% F1 score.
Outcome: The proposed grammar achieves a 95.82% F1-score on the ATIS domain.
Does BERT Rediscover a Classical NLP Pipeline? (2022.coling-1)

Copied to clipboard

Challenge: Existing theories of BERT's structure lack conclusive empirical support . however, there is scepticism about the premises of probing itself .
Approach: They propose a new probe called GridLoc that can take into account token positions, training rounds, and random seeds.
Outcome: The proposed probe detects other, stronger regularities suggesting appeals to layer depth may not be the preferable mode of explanation for BERT’s inner workings.
Tiny Budgets, Big Gains: Parameter Placement Strategy in Parameter Super-Efficient Fine-Tuning (2025.emnlp-main)

Copied to clipboard

Challenge: Existing methods such as LoRA and VeRA use memory-efficient methods to fine-tune large language models.
Approach: They propose a method that uses only 1–5% of the standard LoRA parameters and achieves state-of-the-art performance across a wide range of tasks.
Outcome: The proposed method achieves state-of-the-art performance across a wide range of tasks using only 1–5% of the standard LoRA parameters.
Bringing the State-of-the-Art to Customers: A Neural Agent Assistant Framework for Customer Service Support (2022.emnlp-industry)

Copied to clipboard

Challenge: Creating agent assistants that can help improve customer service support requires inputs from industry users and their customers as well as knowledge of state-of-the-art natural language processing (NLP) technology.
Approach: They propose to combine expertise from academia and industry to build task/domain-specific Neural Agent Assistants with three high-level components for: (1) Intent Identification, (2) Context Retrieval, and (3) Response Generation.
Outcome: The proposed framework is based on three case studies of industry partners who successfully adapt the framework to their unique challenges.
Llama See, Llama Do: A Mechanistic Perspective on Contextual Entrainment and Distraction in LLMs (2025.acl-long)

Copied to clipboard

Challenge: *contextual entrainment* occurs across a wide range of language models (LMs) and prompt settings.
Approach: They hypothesize that there is a circuit of attention heads that corresponds to the phenomenon *contextual entrainment* . when they "turn off" these heads, the effect of contextual entraining is significantly attenuated.
Outcome: The proposed method shows that LMs assign higher logits to tokens that have previously appeared in the context prompt, even for random tokens.
Temporal Histories of Epidemic Events (THEE): A Case Study in Temporal Annotation for Public Health (2020.lrec-1)

Copied to clipboard

Challenge: Current EBS estimates the occurrence time of events based on coarse metadata such as document publication time.
Approach: They propose a temporal annotation standard THEE-TimeML and a corpus TheeBank . they document the corpus annotation process and demonstrate the immediate benefit .
Outcome: The proposed standards are based on the existing timeML and the corpus TheeBank . the proposed standards demonstrate the immediate benefit to public health applications .
ConTempo: A Unified Temporally Contrastive Framework for Temporal Relation Extraction (2024.findings-acl)

Copied to clipboard

Challenge: Temporal relation extraction (TRE) is a task of classifying temporal relations between events conveyed in narratives.
Approach: They propose a Temporally Contrastive learning model that increases the model’s awareness of the meaning of temporal relations by leveraging their symmetric or antisymmetric properties.
Outcome: The proposed model improves the model's representation of meaning of temporal relations and its ability to integrate with the underlying temporal calculus.
Sheaf Discovery with Joint Computation Graph Pruning and Flexible Granularity (2025.emnlp-main)

Copied to clipboard

Challenge: Experimental results show that DiscoGP extracts sheaves that preserve 93-100% of a model’s performance while comprising only 1-7% of the original weights and connections.
Approach: They propose a framework for extracting self-contained modular units within neural language models (LMs) they use a gradient-based pruning algorithm to prune the original LM to a sparse skeleton .
Outcome: The proposed framework preserves 93-100% of the original model's performance while preserving only 1-7% of the model''s original weights and connections.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations