Papers by Meng-Fen Chiang
Evidence-guided Inference for Neutralized Zero-shot Transfer (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing knowledge transfer frameworks that use label skewness to neutralize biased language are costly and impractical when it comes to scarcely labeled data. |
| Approach: | They propose a neutralized Knowledge Transfer framework to equip pre-trained language models with neutralized transferability. |
| Outcome: | The proposed framework shows that it can be used to train pre-trained models with neutralized transferability . it is compared with baselines with a zero-shot cross-domain transfer setting . |
S-RAG: A Novel Audit Framework for Detecting Unauthorized Use of Personal Data in RAG Systems (2025.acl-long)
Copied to clipboard
| Challenge: | Retrieval-Augmented Generation (RAG) systems rely on external data for accurate and context-specific responses. |
| Approach: | They propose a framework that enables users to determine whether their textual data has been utilized in RAG systems even in black-box settings with no prior system knowledge. |
| Outcome: | The proposed framework achieves an improvement in Accuracy by 19.9% while maintaining strong performance under adversarial defenses. |
Disentangling Reasoning Logic to Resolve Explicit Knowledge Conflicts (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to resolve explicit knowledge conflicts are based on semantic decoding and auxiliary embedding. |
| Approach: | They propose a framework that adjudicates conflicts by structuring the underlying logic. |
| Outcome: | Experiments show that the proposed framework improves on existing models. |
Faithful Persona Steering under Incongruity via Dual-Stream Refinement (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for personalization in large language models often neglect incongruity of human personas . empirical evaluations show that QuirkyMind mitigates drift while preserving "quirks" |
| Approach: | a framework disentangles identity definition from its expression by anchoring traits in a dual-stream latent state. |
| Outcome: | a framework disentangles identity definition from its expression, and it mitigates drift without erasing authentic incongruities. |
Evo-PI: Aligning Medical Reasoning via Evolving Principle-Guided Supervision (2026.acl-long)
Copied to clipboard
| Challenge: | Existing models with static prompts, rules, or reward models are constrained by static supervision, which often fails to shape the underlying reasoning process, leading to brittle generalization and performance saturation in complex decision-making tasks. |
| Approach: | They propose a principle-centric learning framework that treats reasoning principles as explicit, language-based supervision signals that can be generated, evaluated, and iteratively evolved. |
| Outcome: | The proposed framework treats reasoning principles as explicit, language-based supervision signals that can be generated, evaluated, and iteratively evolved. |