Papers by Tao Han

30 papers
FOLIO: Natural Language Reasoning with First-Order Logic (2024.emnlp-main)

Copied to clipboard

Challenge: Existing benchmarks for logical reasoning in large language models lack language naturalness or limited complexity.
Approach: They propose to use first-order logic annotations to evaluate logical reasoning capabilities of large language models.
Outcome: The proposed dataset evaluates the FOL reasoning ability of supervised fine-tuning on medium-sized language models.
Rehearsal-free Continual Language Learning via Efficient Parameter Isolation (2023.acl-long)

Copied to clipboard

Challenge: Existing methods for learning continual tasks do not cache history data, which makes the problem more challenging.
Approach: They propose a method that allocates a small portion of private parameters and learns them with a shared pre-trained model.
Outcome: The proposed method is comparable to existing methods and comparable to those using historical data.
Mitigating Lost in Multi-turn Conversation via Curriculum RL with Verifiable Accuracy and Abstention Rewards (2026.acl-long)

Copied to clipboard

Challenge: Large Language Models exhibit strong capabilities in single-turn instruction following but suffer from Lost-in-Conversation (LiC) when instructions are revealed progressively in multi-turn settings, models get "Lost in Conversation"
Approach: They propose a framework that encourages models to generate correct answers and judge solvability in multi-turn conversations.
Outcome: The proposed framework improves models' ability to balance problem-solving with abstention . it reduces premature answering behaviors that cause lost-in-conversation (LiC)
DeepSynth-Eval: Objectively Evaluating Information Consolidation in Deep Survey Writing (2026.findings-acl)

Copied to clipboard

Challenge: Large Language Models (LLMs) are evolving towards autonomous agents . retrieval capabilities are well-benchmarked, but post-retrieval synthesis is under-evaluated due to open-ended writing.
Approach: They propose a benchmark to evaluate information consolidation capabilities using survey papers as gold standards.
Outcome: The proposed benchmark analyzes the post-retrieval synthesis stage of large language models . it leverages high-quality survey papers as gold standards and reverse-engineers research requests . the proposed benchmark outperforms single-turn generation and reduces hallucinations .
BPP-Search: Enhancing Tree of Thought Reasoning for Mathematical Modeling Problem Solving (2025.acl-long)

Copied to clipboard

Challenge: Existing datasets in operations research domain lack detailed annotations of the modeling process, focusing only on objective values.
Approach: They propose an annotation-based tree-of-thought tree-based reasoning algorithm that integrates reinforcement learning into a tree- of-though.
Outcome: The proposed algorithm outperforms state-of-the-art methods on StructuredOR, NL4OPT, and MAMO-ComplexLP datasets.
Orthogonal Subspace Learning for Language Model Continual Learning (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing methods for continual learning in language models suffer catastrophic forgetting when learning sequential tasks.
Approach: They propose an orthogonal low-rank adaptation approach for continual learning in language models that uses orthogons to learn sequentially.
Outcome: The proposed approach outperforms state-of-the-art methods on continual learning benchmarks and preserves generalization ability of LLMs on unseen tasks.
“I’ve Decided to Leak”: Probing Internals Behind Prompt Leakage Intents (2025.emnlp-main)

Copied to clipboard

Challenge: Large language models (LLMs) exhibit prompt leakage vulnerabilities, raising intellectual property and confidentiality concerns.
Approach: They use probing techniques to capture LLMs’ intent-related internal representations and show that they internalize prompt leakage intents in their hidden states before generating tokens.
Outcome: The proposed probes achieve 90%+ AUROC across all tested models, even when applied to new system prompts and attacks.
LASA: Language-Agnostic Semantic Alignment at the Semantic Bottleneck for LLM Safety (2026.acl-long)

Copied to clipboard

Challenge: Large language models (LLMs) have demonstrated better safety performance in high-resource languages than in low-resourced languages.
Approach: They propose language-agnostic semantic alignment (LASA) which anchors safety alignment directly in semantic bottlenecks.
Outcome: The proposed approach significantly improves safety across all languages: average attack success rate drops from 24.7% to 2.8% on LLaMA-3.1-8B-Instruct and remains within 3–4% across Qwen2.5 and Qwend3 Instruct models (7B–32B).
TencentPretrain: A Scalable and Flexible Toolkit for Pre-training Models of Different Modalities (2023.acl-demo)

Copied to clipboard

Challenge: Several pre-training models of different modalities are showing a rising trend of homogeneity in their model structures.
Approach: They propose a toolkit that supports pre-training models of different modalities.
Outcome: The proposed toolkit can match the performance of the original implementations on text, vision, and audio benchmarks.
RouteMoA: Dynamic Routing without Pre-Inference Boosts Efficient Mixture-of-Agents (2026.acl-long)

Copied to clipboard

Challenge: Existing methods for mixing-of-agents (MoA) lack model selection criteria and struggle with large model pools.
Approach: They propose a mixture-of-agents framework with dynamic routing that uses a lightweight scorer to perform initial screening and refines the model scores through self- and cross-assessment.
Outcome: The proposed framework outperforms existing methods for large model pools and tasks . it reduces cost by 89.8% and latency by 63.6% in the large-scale model pool.
VideoPro: Adaptive Program Reasoning for Long Video Understanding (2026.acl-long)

Copied to clipboard

Challenge: Existing methods for understanding long videos are limited due to the sparsity of visual evidence relevant to a given query.
Approach: They propose a framework that enables VideoLLMs to reason over long videos and refine their predictions through executable programs.
Outcome: The proposed framework outperforms existing methods across long-video understanding benchmarks.
Arxiv Copilot: A Self-Evolving and Efficient LLM System for Personalized Academic Assistance (2024.emnlp-demo)

Copied to clipboard

Challenge: Existing solutions for document QA fail to provide personalized and up-to-date information efficiently.
Approach: They propose to deploy a self-evolving, efficient LLM system that can offer personalized research services, maintaining a real-time updated database.
Outcome: The proposed system saves 69.92% of time after efficient deployment.
BCL: Bayesian In-Context Learning Framework for Information Extraction (2026.findings-acl)

Copied to clipboard

Challenge: Existing information extraction (IE) tasks rely on in-context learning with large language models.
Approach: They propose a Bayesian-based in-context learning framework that refines label representations across IE tasks using particle filtering and Bayes updates.
Outcome: The proposed framework improves performance over existing methods (up to 30%) it underperforms one-shot prompting by a substantial margin on NER tasks and CodeIE fails on RE tasks with near-zero micro-F1.
LoRE-Merging: Exploring Low-Rank Estimation For Large Language Model Merging (2025.findings-emnlp)

Copied to clipboard

Challenge: a framework for model merging is proposed without additional training . task vectors from fine-tuned models exhibit a limited number of dominant singular values .
Approach: They propose a framework for model merging based on low-rank estimation of task vectors without access to the base model.
Outcome: The proposed framework improves models without additional training without additional inputs.
InfiMM: Advancing Multimodal Understanding with an Open-Sourced Visual Language Model (2024.findings-acl)

Copied to clipboard

Challenge: InfiMM is a multimodal large language model that adapts to complex vision-language tasks.
Approach: They present a Multimodal Large Language Model that adapts to intricate vision-language tasks using large-scale training data and comprehensive training strategies.
Outcome: Empirical evaluations across a variety of benchmarks underscore InfiMM’s remarkable capability in multimodal understanding.
RACE: Retrieval-augmented Commit Message Generation (2022.emnlp-main)

Copied to clipboard

Challenge: Existing approaches to automatically generate commit messages are repetitive or redundant.
Approach: They propose a retrieval-augmented neural commit message generation method which treats the retrieved similar commit as an exemplar and leverages it to generate an accurate commit message.
Outcome: The proposed method outperforms baselines on a large dataset with five programming languages and can boost existing Seq2Seq models in commit message generation.
Finding the Evidence: Localization-aware Answer Prediction for Text Visual Question Answering (2020.coling-main)

Copied to clipboard

Challenge: Existing text VQA systems generate an answer by selecting from optical character recognition (OCR) texts or a fixed vocabulary.
Approach: They propose a localization-aware answer prediction network that generates the answer and predicts a bounding box as evidence of the generated answer.
Outcome: The proposed network outperforms existing methods on three benchmark datasets for the text VQA task by a noticeable margin.
Robust and Scalable Model Editing for Large Language Models (2024.lrec-main)

Copied to clipboard

Challenge: Existing methods that ignore contextual knowledge fail to reliably fall back to parametric knowledge when presented with irrelevant context.
Approach: They propose to use contextual knowledge to update and correct LLMs' knowledge by in-context editing instead of retraining.
Outcome: The proposed method outperforms current state-of-the-art methods by a large margin on a dataset that contains irrelevant questions.
Multi-stage Distillation Framework for Cross-Lingual Semantic Similarity Matching (2022.findings-naacl)

Copied to clipboard

Challenge: Existing studies have shown that cross-lingual knowledge distillation can improve the performance of pre-trained models for cross-linguistic similarity matching tasks.
Approach: They propose a multi-stage distillation framework for constructing a small-size but high-performance cross-lingual model using contrastive learning, bottleneck, and parameter recurrent strategies.
Outcome: The proposed model can compress the size of XLM-R and MiniLM by more than 50% while the performance is only reduced by about 1%.
Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework (2026.findings-acl)

Copied to clipboard

Challenge: Existing studies on how SAEs derive most fine-grained latent features for safety remain unexplored.
Approach: They propose a framework for interpreting SAE features in safety-critical domains . they train a suite of SAEs with human-readable explanations and systematic evaluations based on pornography, politics, violence, and terror .
Outcome: The proposed framework reduces interpretation cost by 55% and improves safety-critical features.
PromptIntern: Saving Inference Costs by Internalizing Recurrent Prompt during Large Language Model Fine-tuning (2024.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in fine-tuning large language models have greatly enhanced their usage in domain-specific tasks.
Approach: They propose a method which internalizes prompt knowledge during model fine-tuning to achieve efficient inference and save costs.
Outcome: The proposed approach reduces input tokens by 90%, accelerates inference by 4.2 times, and reduces monetary inference costs by 88.3%.
MM-Doc-R1: Training Agents for Long Document Visual Question Answering through Multi-turn Reinforcement Learning (2026.findings-acl)

Copied to clipboard

Challenge: Existing work on long document visual question answering is based on Retrieval-Augmented Generation (RAG) where textual or visual content is encoded into embeddings and relevance is determined by similarity scores with respect to the original query.
Approach: They propose a framework that employs an agentic, vision-aware workflow to address long document visual question answering through iterative information discovery and synthesis.
Outcome: The proposed framework outperforms existing RL systems by 10.4% on the MMLongbench-Doc benchmark and demonstrates superior training performance over GRPO.
Revisiting the Reliability of Language Models in Instruction-Following (2026.acl-long)

Copied to clipboard

Challenge: Several benchmarks have been proposed to measure instruction-following accuracy, but these scores do not translate to reliable services in real-world use.
Approach: They propose a new metric reliable@k and develop an automated pipeline to generate cousin prompts.
Outcome: The proposed model can be instantiated with cousin prompts and generates high-quality cousin prompt data.
RoCoIns: Enhancing Robustness of Large Language Models through Code-Style Instructions (2024.lrec-main)

Copied to clipboard

Challenge: Large Language Models (LLMs) have shown remarkable capabilities in following human instructions and solving NLU tasks.
Approach: They propose to use code style instructions to replace typically natural language instructions to provide more precise instructions and strengthen the robustness of LLMs.
Outcome: The proposed method outperforms natural language models on eight robustness datasets and achieves an improvement of 5.68% in test set accuracy and a reduction of 5.66 points in Attack Success Rate (ASR).
LAD-RAG: Layout-aware Dynamic RAG for Visually-Rich Document Understanding (2026.acl-long)

Copied to clipboard

Challenge: Conventional retrieval-augmented generation (RAG) methods encode content in isolated chunks during ingestion, losing structural and cross-page dependencies, and retrieve a fixed number of pages at inference.
Approach: They propose a Layout-Aware Dynamic RAG framework that encodes content in isolated chunks during ingestion and retrieves a fixed number of pages at inference.
Outcome: Experiments on MMLongBench-Doc, LongDocURL, DUDE, and MP-DoxVQA show that LAD-RAG improves retrieval, achieving over 90% perfect recall on average without any top-k tuning, and outperforming baseline retrievers by up to 20% in recall at comparable noise levels.
ELBA-Bench: An Efficient Learning Backdoor Attacks Benchmark for Large Language Models (2025.acl-long)

Copied to clipboard

Challenge: Existing backdoor models are limited in coverage of attack, system integrity and backdoor alignment . ELBA-Bench provides over 1300 experiments encompassing 12 attack methods, 18 datasets, and 12 LLMs.
Approach: They propose a framework that allows attackers to inject backdoor through parameter efficient fine-tuning or without fine-uning techniques.
Outcome: ELBA-Bench provides over 1300 experiments encompassing 12 attack methods, 18 datasets, and 12 LLMs.
A Structure-Aware Generative Adversarial Network for Bilingual Lexicon Induction (2023.findings-emnlp)

Copied to clipboard

Challenge: Bilingual lexicon induction (BLI) is the task of inducing word translations with a learned mapping function that aligns monolingual word embedding spaces in two different languages.
Approach: They propose a model that explicitly captures multiple topological structure information to achieve accurate bilingual lexicon induction.
Outcome: The proposed model captures multiple topological structure information to achieve accurate BLI on a public dataset.
Mitigating Coordinate Prediction Bias from Positional Encoding Failures (2026.findings-acl)

Copied to clipboard

Challenge: Multimodal Large Language Models (MLLMs) excel at general vision-language tasks, but precise coordinate prediction remains a challenge.
Approach: They propose a training-free, inference-time correction method to correct VPEs . they isolate position-unconditioned tendencies by shuffling VPE and use it to steer digit decoding .
Outcome: The proposed method is training-free, inference-time correction method . it effectively rectifies coordinate drift, yielding consistent improvements without retraining .
FormLM: Recommending Creation Ideas for Online Forms by Modelling Semantic and Structural Information (2022.emnlp-main)

Copied to clipboard

Challenge: FormLM is a pre-trained language model for creating semi-structured forms where questions and descriptions are organized by predefined structures.
Approach: They propose to enhance pre-trained language model with form structural information to model online forms and recommend form creation ideas.
Outcome: The proposed model outperforms general-purpose language models on all tasks, with an improvement by 4.71 on Question Recommendation and 10.6 on Block Type Suggestion in terms of ROUGE-1 and Macro-F1, respectively.
ProSparse: Introducing and Enhancing Intrinsic Activation Sparsity within Large Language Models (2025.coling-main)

Copied to clipboard

Challenge: Activation sparsity is a promising paradigm for accelerating model inference . few large language models achieve high activation spar and comparable performance .
Approach: They propose a method to achieve activation sparsity and acceleration in large language models . they introduce ReLU activation and adopt progressive sparse regularization .
Outcome: The proposed method achieves high activation sparsity and comparable model performance.

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