Papers by Jun Han

24 papers
Fast and Effective On-Policy Distillation from Reasoning Prefixes (2026.findings-acl)

Copied to clipboard

Challenge: On-policy distillation (OPD) requires expensive on-the-fly sampling of the student policy during training, which substantially increases training cost.
Approach: They propose to use on-policy distillation to sample trajectories from student model . they propose to terminate the sampling early during distillation .
Outcome: The proposed method matches the performance of full OPD in long reasoning outputs while reducing training FLOP by 2x–40x.
Direct Behavior Optimization: Unlocking the Potential of Lightweight LLMs (2025.findings-acl)

Copied to clipboard

Challenge: Existing prompt optimization methods rely on extensive manual effort or meta-cognitive abilities, making them less effective for LwLLMs.
Approach: They propose a direct behavior optimization parameter that transforms the optimization of complex prompts into discrete, quantifiable execution sequences using a gradient-free Monte Carlo Tree Search.
Outcome: The proposed method outperforms current prompt optimization methods on seven challenging tasks where state-of-the-art LLMs excel but LwLLMs generally underperform.
SAGE: Steerable Agentic Data Generation for Deep Search with Execution Feedback (2026.findings-eacl)

Copied to clipboard

Challenge: High-quality, complex question-answer pairs are pivotal for training and evaluating capable deep search agents.
Approach: They propose a pipeline that generates high-quality, difficulty-controlled deep search question-answer pairs for a given corpus and a target difficulty level.
Outcome: The proposed pipeline generates high-quality, difficulty-controlled deep search question-answer pairs for a given corpus and a target difficulty level.
Layer-wise Fusion with Modality Independence Modeling for Multi-modal Emotion Recognition (2023.acl-long)

Copied to clipboard

Challenge: Existing studies focus on developing models that exploit the unification of multiple modalities.
Approach: They propose to maintain modality independence by using a multi-modal transformer model that fuses all modalities.
Outcome: The proposed model outperforms state-of-the-art models in multi-modal emotion recognition.
Direct Value Optimization: Improving Chain-of-Thought Reasoning in LLMs with Refined Values (2025.emnlp-main)

Copied to clipboard

Challenge: Existing offline preference optimization methods rely on preference labels to optimize large language models.
Approach: They propose an offline method for enhancing large language models in reasoning tasks that utilizes value signals at individual reasoning steps.
Outcome: The proposed framework outperforms offline preference optimization techniques by 4% to 6% on math reasoning, commonsense reasoning, and coding tasks.
DisCal: Distribution-Aware Calibration for Mathematical Reasoning Under Character-Level Noisy Inputs (2026.acl-long)

Copied to clipboard

Challenge: Existing methods for calibration of large reasoning models (LRMs) focus on clean inputs, leaving noise unexplored.
Approach: They propose a confidence calibration framework for character-level noisy inputs that extracts uncertainty signals from both the empirical answer distribution and the model’s predictive distribution and integrates them via a learned calibrator.
Outcome: Experiments on multiple mathematical reasoning benchmarks show that DisCal outperforms existing calibration methods under noisy inputs, reducing expected calibration error (ECE) by up to 39.21% and improving Area Under the Receiver Operating Characteristic Curve (AUROC) by 31.44%.
In Prospect and Retrospect: Reflective Memory Management for Long-term Personalized Dialogue Agents (2025.acl-long)

Copied to clipboard

Challenge: Existing approaches to long-term dialogue memory management fail to capture the natural semantic structure of conversations, leading to fragmented and incomplete representations.
Approach: They propose a mechanism that integrates forward- and backward-looking reflections into a personalized memory bank for effective future retrieval.
Outcome: The proposed mechanism outperforms state-of-the-art benchmarks on a long-term dialogue memory model.
IntactKV: Improving Large Language Model Quantization by Keeping Pivot Tokens Intact (2024.findings-acl)

Copied to clipboard

Challenge: Existing quantization methods are compromising performance of large language models (LLMs) despite their high computational intensity, LLMs are still demanding intensive computation.
Approach: They propose to generate the KV cache of pivot tokens losslessly from the full-precision model.
Outcome: The proposed method generates the KV cache of pivot tokens losslessly from the full-precision model with no extra inference overhead.
Test-Time Code-Switching for Cross-lingual Aspect Sentiment Triplet Extraction (2025.naacl-long)

Copied to clipboard

Challenge: Aspect Sentiment Triplet Extraction (ASTE) is a thriving research area . current code-switching methods suffer from term boundary detection issues and out-of-dictionary problems.
Approach: They propose a test-time code-switching framework which bridges the gap between bilingual training and monolingual test- time prediction.
Outcome: The proposed framework achieves an average improvement of 3.7% on four cross-lingual datasets.
An MRC Framework for Semantic Role Labeling (2022.coling-1)

Copied to clipboard

Challenge: Existing work on semantic role labels ignores the semantic connection between the two tasks . et al. (2010) defined two types of semantic roles: core roles and non-core roles.
Approach: They propose to use machine reading comprehension to bridge the gap between these two tasks . they formalize predicate disambiguation as multiple-choice machine reading understanding .
Outcome: The proposed framework achieves state-of-the-art or comparable results to previous work . it uses the descriptions of candidate senses of a given predicate as options to select the correct sense .
MSIA: Adaptive Medical Multimodal Multi-turn Semantic Jailbreak (2026.findings-acl)

Copied to clipboard

Challenge: Medical large language models exhibit high domain specificity and condensed semantics, making them vulnerable to diagnostic errors in real-world clinical settings.
Approach: They propose a framework for modeling and inducing multi-turn medical semantic jailbreaks in clinical dialogues.
Outcome: Experiments on chest X-ray-based multimodal medical dialogues show that MSIA outperforms existing jailbreak methods with an average success rate of 76.67%.
PhysPRM: A Generative Process Reward Model with Fine-grained Diagnosis for Physics Problem Solving (2026.findings-acl)

Copied to clipboard

Challenge: Existing Large Language Models (LLMs) struggle with physics problem solving due to difficulties in decoding implicit constraints and maintaining physical consistency.
Approach: They propose a Generative PRM that treats evaluation as a generative task . it produces fine-grained diagnoses comprising critiques, final judgments, and specific error types .
Outcome: The proposed model improves performance across seven benchmarks in Best-of-N and critique refinement strategies.
FloorPlan-LLaMa: Aligning Architects’ Feedback and Domain Knowledge in Architectural Floor Plan Generation (2025.acl-long)

Copied to clipboard

Challenge: Existing evaluation methods for floor plan generation rely on statistical metrics like FID, GED, and PSNR, which fail to evaluate using domain knowledge.
Approach: They propose to use a first floor plan dataset to train a floor plan generation model based on a multi-dimensional preference score and a textual analysis to integrate architects’ professional expertise and preferences.
Outcome: The proposed model outperforms baseline models in text-conditional and class-condition tasks and is more rational and aligns better with human preferences.
BMInf: An Efficient Toolkit for Big Model Inference and Tuning (2022.acl-demo)

Copied to clipboard

Challenge: Recent years, pre-trained language models (PLMs) have achieved promising results on various NLP tasks.
Approach: They propose an open-source toolkit for big model inference and tuning which can support big model tuning at extremely low computation cost.
Outcome: The proposed toolkit can support big model inference and tuning at extremely low computation cost.
An Adaptive Prompt Generation Framework for Task-oriented Dialogue System (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing black-box large language models (LLMs) have excellent performance in task-oriented dialogue (TOD) tasks, but obtaining suitable prompts for specific tasks is challenging.
Approach: They propose a black-box large language model that generates domain and slot information in the belief state, which serves as prior knowledge for subsequent prompt generation.
Outcome: The proposed framework outperforms existing prompting methods on the MultiWOZ 2.0 dataset.
Ask Question First for Enhancing Lifelong Language Learning (2022.coling-1)

Copied to clipboard

Challenge: Existing approaches to stream learning NLP tasks suffer from catastrophic forgetting and are exacerbated when the previous task’s pseudo data is insufficient.
Approach: They propose to use a new data format to train pseudo questions of previous tasks to stream learning NLP tasks while retaining knowledge of previous ones.
Outcome: The proposed model is more robust to sufficient and insufficient pseudo-data when the task boundary is both clear and unclear.
Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia (2025.acl-long)

Copied to clipboard

Samuel Cahyawijaya, Holy Lovenia, Joel Ruben Antony Moniz, Tack Hwa Wong, Mohammad Rifqi Farhansyah, Thant Thiri Maung, Frederikus Hudi, David Anugraha, Muhammad Ravi Shulthan Habibi, Muhammad Reza Qorib, Amit Agarwal, Joseph Marvin Imperial, Hitesh Laxmichand Patel, Vicky Feliren, Bahrul Ilmi Nasution, Manuel Antonio Rufino, Genta Indra Winata, Rian Adam Rajagede, Carlos Rafael Catalan, Mohamed Fazli Mohamed Imam, Priyaranjan Pattnayak, Salsabila Zahirah Pranida, Kevin Pratama, Yeshil Bangera, Adisai Na-Thalang, Patricia Nicole Monderin, Yueqi Song, Christian Simon, Lynnette Hui Xian Ng, Richardy Lobo Sapan, Taki Hasan Rafi, Bin Wang, null Supryadi, Kanyakorn Veerakanjana, Piyalitt Ittichaiwong, Matthew Theodore Roque, Karissa Vincentio, Takdanai Kreangphet, Phakphum Artkaew, Kadek Hendrawan Palgunadi, Yanzhi Yu, Rochana Prih Hastuti, William Nixon, Mithil Bangera, Adrian Xuan Wei Lim, Aye Hninn Khine, Hanif Muhammad Zhafran, Teddy Ferdinan, Audra Aurora Izzani, Ayushman Singh, Evan Evan, Jauza Akbar Krito, Michael Anugraha, Fenal Ashokbhai Ilasariya, Haochen Li, John Amadeo Daniswara, Filbert Aurelian Tjiaranata, Eryawan Presma Yulianrifat, Can Udomcharoenchaikit, Fadil Risdian Ansori, Mahardika Krisna Ihsani, Giang Nguyen, Anab Maulana Barik, Dan John Velasco, Rifo Ahmad Genadi, Saptarshi Saha, Chengwei Wei, Isaiah Edri W. Flores, Kenneth Chen Ko Han, Anjela Gail D. Santos, Wan Shen Lim, Kaung Si Phyo, Tim Santos, Meisyarah Dwiastuti, Jiayun Luo, Jan Christian Blaise Cruz, Ming Shan Hee, Ikhlasul Akmal Hanif, M.Alif Al Hakim, Muhammad Rizky Sya’ban, Kun Kerdthaisong, Lester James Validad Miranda, Fajri Koto, Tirana Noor Fatyanosa, Alham Fikri Aji, Jostin Jerico Rosal, Jun Kevin, Robert Wijaya, Onno P. Kampman, Ruochen Zhang, Börje F. Karlsson, Peerat Limkonchotiwat
Challenge: Southeast Asia is underrepresented in vision-language research . SEA-VL is an open-source initiative dedicated to developing culturally relevant datasets for SEA languages.
Approach: They propose to use crowdsourced, automated image crawling and synthetic image generation to develop culturally relevant datasets for SEA languages.
Outcome: The proposed datasets capture SEA cultural nuances and contexts better than existing datasets.
Decomposing Complex Questions Makes Multi-Hop QA Easier and More Interpretable (2021.findings-emnlp)

Copied to clipboard

Challenge: Multi-hop QA requires the machine to answer complex questions through finding multiple clues and reasoning, and provide explanatory evidence to demonstrate the reasoning process.
Approach: They propose a three-stage framework based on complex question decomposition that decomposes the complex question, then reads the sub-questions and then performs numerical comparison to get the final answer.
Outcome: The proposed framework achieves state-of-the-art in the 2WikiMultiHopQA dataset, with a winning joint F1 score of 53.58 on the leaderboard.
The UIR Uncertainty Corpus for Chinese: Annotating Chinese Microblog Corpus for Uncertainty Identification from Social Media (L18-1)

Copied to clipboard

Challenge: Uncertainty identification is an important semantic processing task, critical to the quality of information in terms of factuality in many NLP techniques and applications.
Approach: They propose to annotate Chinese microblogs with an open uncertainty corpus . they propose to use contextual uncertain semantics rather than traditional cue-phrases to identify uncertainty .
Outcome: The proposed corpus can be used to identify uncertainty in social media texts.
Efficient Paths and Dense Rewards: Probabilistic Flow Reasoning for Large Language Models (2026.acl-long)

Copied to clipboard

Challenge: Existing approaches to mitigate inference inefficiency and optimization difficulty are fragmented and constrained by inherent trade-offs.
Approach: They propose a framework that reconceptualizes discrete reasoning steps as a continuous probabilistic flow, quantifying the contribution of each step toward the ground-truth answer.
Outcome: The proposed framework achieves a superior balance between inference efficiency and reasoning performance on challenging benchmarks.
Compiling Activation Steering into Weights via Null-Space Constraints for Stealthy Backdoors (2026.acl-long)

Copied to clipboard

Challenge: Existing methods to inject safety-aligned large language models rely on token-level mappings, which do not guarantee sustained harmful output.
Approach: They propose a method that directly modifies model weights to map a trigger to an attacker-specified response.
Outcome: The proposed method achieves high triggered attack success while maintaining non-triggered safety and general utility.
SafeConf: A Confidence-Calibrated Safety Self-Evaluation Method for Large Language Models (2025.findings-emnlp)

Copied to clipboard

Challenge: Large language models (LLMs) have many advantages but they also pose significant safety risks.
Approach: They propose a method to enhance the safety self-evaluation capability of LLMs . they perform semantic mutations on the original safety evaluation questions .
Outcome: The proposed method improves safety self-evaluation accuracy by 5.86% and 7.79% over baseline methods on Chinese and English datasets.
Conjoin after Decompose: Improving Few-Shot Performance of Named Entity Recognition (2024.lrec-main)

Copied to clipboard

Challenge: Existing prompt-based NER models fail to detect entity boundaries, causing performance degradation.
Approach: They propose a model which consists of a BART encoder and a parabiotic decoder and propose ' boundary expansion strategy' to enhance the model's capability in entity type classification.
Outcome: The proposed model can achieve significant performance gains over state-of-the-art models.
SepSeq: A Training-Free Framework for Long Numerical Sequence Processing in LLMs (2026.findings-acl)

Copied to clipboard

Challenge: Existing large-scale large-context models suffer from performance degradation when processing long numerical sequences.
Approach: They propose a framework to mitigate attention dispersion by strategically inserting separator tokens into the model to recalibrat attention to local segments while preserving global context.
Outcome: The proposed framework improves accuracy and reduces inference token consumption by 16.4% on 9 widely-adopted LLMs.

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