Papers by Zhiyu Ni

4 papers
Dictionary Guided Sparse Logit Editing for Reliable Jailbreak Attacks (2026.findings-acl)

Copied to clipboard

Challenge: Existing methods to optimize large language models suffer from high computational costs and produce uninterpretable, high-perplexity inputs.
Approach: They propose a sparse index-based intervention that bypasses guardrails via sparser logit editing.
Outcome: The proposed method bypasses guardrails by modifying pre-softmax logits without gradients or auxiliary models.
Defending LLMs against Jailbreak Attacks via Template-Based ICL with a Defensive Suffix (2026.findings-acl)

Copied to clipboard

Challenge: State-of-the-art large language models (LLMs) are vulnerable to jailbreak attacks, such as GCG and AutoDAN.
Approach: They propose to take the advances of online In-Context Learning and an offline defensive suffix and optimize them using an iterative algorithm and an online stochastic random search to identify the most effective ICL demonstrations.
Outcome: The proposed method reduces attack success rate to nearly *0% while maintaining the model’s utility on benign tasks and incurring only *negligible* computational overhead.
LMR-BENCH: Evaluating LLM Agent’s Ability on Reproducing Language Modeling Research (2025.emnlp-main)

Copied to clipboard

Challenge: Large language model (LLM) agents have demonstrated remarkable potential in advancing scientific discovery, but their capability in reproducing code from research papers remains underexplored.
Approach: They propose to evaluate LLM agents' ability to reproduce scientific research papers by analyzing code reproduction tasks from 23 research papers published in top-tier NLP venues.
Outcome: The proposed benchmark systematically evaluates the capability of large language model (LLM) agents on code reproduction from Language Modeling Research.
Defenses Against Prompt Attacks Learn Surface Heuristics (2026.acl-long)

Copied to clipboard

Challenge: Large language models (LLMs) are increasingly deployed in security-sensitive applications . recent defenses rely on supervised fine-tuning with benign and malicious labels . position bias arises when benign content placed later in a prompt is rejected at much higher rates .
Approach: They analyze three recurring shortcut behaviors induced by supervised fine-tuning . position bias arises when benign content placed later in a prompt is rejected . token trigger bias occurs when strings common in attack data raise rejection probability .
Outcome: The proposed model overrides intended logic when adversarial instructions appear . the proposed model has low rejection rates but narrow correlations in defense data .

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