Papers by Ziheng Li

10 papers
TinyScientist: An Interactive, Extensible, and Controllable Framework for Building Research Agents (2025.emnlp-demos)

Copied to clipboard

Challenge: Existing research systems often design and use agentic workflows to perform research tasks such as ideation, scientific coding, review writing, and tree-based search.
Approach: They propose an open-source codebase, an interactive web demonstration, and a PyPI Python package to make state-of-the-art auto-research pipelines broadly accessible to every researcher and developer.
Outcome: The proposed framework adapts easily to new tools and supports iterative growth.
DISC: Plug-and-Play Decoding Intervention with Similarity of Characters for Chinese Spelling Check (2025.acl-long)

Copied to clipboard

Challenge: Chinese spelling check (CSC) tasks require that incorrect characters are usually similar to the correct ones in either phonetics or glyph.
Approach: They propose a plug-and-play decoding intervention with similarity of characters module for Chinese spelling check (CSC) they propose to incorporate phonetic and glyph similarities only during the inference phase.
Outcome: The proposed method significantly improves Chinese spelling check models on benchmarks and on benchmark datasets.
Amphista: Bi-directional Multi-head Decoding for Accelerating LLM Inference (2025.naacl-long)

Copied to clipboard

Challenge: Existing methods such as Medusa lack adequate information interaction between different drafting heads.
Approach: They propose an enhanced speculative decoding framework that builds upon Medusa and integrates a drafting block capable of parallel inference.
Outcome: The proposed framework outperforms Medusa in terms of head accuracy and latency.
AdaLoGN: Adaptive Logic Graph Network for Reasoning-Based Machine Reading Comprehension (2022.acl-long)

Copied to clipboard

Challenge: Existing methods and limitations for machine reading comprehension are insufficient for logical reasoning over text.
Approach: They propose a neural-symbolic approach which passes messages over a graph representing logical relations between text units to predict an answer.
Outcome: The proposed approach outperforms existing methods on ReClor and LogiQA.
Dual-Alignment Pre-training for Cross-lingual Sentence Embedding (2023.acl-long)

Copied to clipboard

Challenge: Recent studies have shown that dual encoder models trained with the sentence-level translation ranking task are effective methods for cross-lingual sentence embedding.
Approach: They propose a dual-alignment pre-training framework that incorporates both sentence-level and token-level alignment.
Outcome: The proposed framework improves cross-lingual sentence embedding on three cross-linguistic benchmarks.
Relation-aware Ensemble Learning for Knowledge Graph Embedding (2023.emnlp-main)

Copied to clipboard

Challenge: Existing methods to explore semantics of knowledge graphs have been proposed to explore these semantics in distinct ways.
Approach: They propose to leverage existing methods in relation-aware manner to learn an ensemble by leveraging existing methods.
Outcome: The proposed method has the same computation cost as general ensemble methods but with much better performance on benchmark datasets.
Decoder-Only LLMs can be Masked Auto-Encoders (2025.acl-short)

Copied to clipboard

Challenge: Modern NLP workflows require different models for generation and embedding tasks.
Approach: They propose a method that transforms an LLM into a Uni-Directional Masked Auto-Encoder.
Outcome: The proposed method achieves state-of-the-art under unsupervised conditions with merely 100 training steps.
Mixture of Small and Large Models for Chinese Spelling Check (2025.acl-long)

Copied to clipboard

Challenge: Chinese Spelling Check (CSC) tasks have been developed to correct spelling errors in given sentences . fine-tuned BERT-based models show excellent performance but suffer from edit pattern overfitting . a novel mixture approach that effectively combines small models and LLMs during beam search decoding phase improves accuracy and fluency of LLM.
Approach: They propose a dynamic mixture approach that effectively combines small models and LLMs during beam search decoding phase.
Outcome: The proposed method significantly boosts error correction capabilities, achieving state-of-the-art results across multiple datasets.
Outcome-Grounded Advantage Reshaping for Fine-Grained Credit Assignment in Mathematical Reasoning (2026.acl-long)

Copied to clipboard

Challenge: Group Relative Policy Optimization (GRPO) uses a coarse-grained credit assignment mechanism that propagates group-level rewards uniformly to to every token in a sequence, neglecting the varying contribution of individual reasoning steps.
Approach: They introduce Outcome-grounded Advantage Reshaping (OAR) which redistributes advantages based on how much each token influences the model’s final answer.
Outcome: Empirical results show that OAR-G outperforms GRPO on a high-fidelity attribution signal and suppresses low-impact tokens while preserving the advantage mass.
EquiBench: Benchmarking Large Language Models’ Reasoning about Program Semantics via Equivalence Checking (2025.emnlp-main)

Copied to clipboard

Challenge: EquiBench is a new benchmark to evaluate large language models' ability to reason about program semantics . Unlike natural language, code is executable.
Approach: They propose a benchmark to evaluate large language models through equivalence checking . EquiBench consists of 2400 program pairs across four languages and six categories .
Outcome: The proposed benchmark consists of 2400 program pairs across four languages and six categories.

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