Skip to yearly menu bar Skip to main content

Workshop

Secure and Trustworthy Large Language Models

Yisen Wang ⋅ Ting Wang ⋅ Jinghui Chen ⋅ Chaowei Xiao ⋅ Jieyu Zhao ⋅ Nanyun (Violet) Peng ⋅ Yulia Tsvetkov ⋅ Anima Anandkumar

Project Page [ Contact: cxiao34@wisc.edu ]

Abstract

Large Language Models (LLMs) have emerged as transformative tools in natural language processing, redefining benchmarks across tasks from machine translation to dialog systems. However, with these advancements come intricate challenges centered around the security, transparency, and ethical dimensions of LLMs. These challenges, ranging from biases and misinformation dissemination to vulnerabilities against sophisticated attacks, have garnered considerable research attention. Our proposed workshop seeks to shine a spotlight on these pivotal issues, focusing on a myriad of topics including, but not limited to, LLM reliability, interpretability, backdoor defenses, and emerging learning paradigms. This assembly aims to bridge gaps between academia and industry, offering a platform for rigorous discussion, collaborative brainstorming, and a showcase of the latest research breakthroughs. Through this endeavor, we aspire to pave a pathway towards more secure, transparent, and ethically-grounded developments in LLMs, underlining the importance of collaborative, cross-disciplinary efforts in the process.

Video

Chat is not available.

Schedule

Timezone: America/Los_Angeles

12:00 AM

Opening remarks

12:10 AM

Invited Talk 1 -Tatsu Hashimoto

Video

12:40 AM

Oral Paper Presentation 1

Video

12:50 AM

Oral Paper Presentation 2

Video

1:00 AM

Invited Talk 2 - Graham Neubig

Video

1:30 AM

Oral Paper Presentation 3

Video

1:40 AM

Oral Paper Presentation 4

Video

1:50 AM

Poster Session A

3:00 AM

Lunch break

4:00 AM

Invited Talk 3 - Bo Li

Video

4:30 AM

Invited Talk 4 - Robin Jia

Video

5:00 AM

Invited Talk 5 - Tom Goldstein

Video

5:30 AM

Invited Talk 6 - Chaowei Xiao

Video

6:00 AM

Invited Talk 7 - Eric Wallace

Video

6:30 AM

Oral Paper Presentation 5

Video

6:45 AM

Oral Paper Presentation 6

Video

7:00 AM

Poster Session B

7:50 AM

Closing Remarks

Group Preference Optimization: Few-Shot Alignment of Large Language Models

Siyan Zhao ⋅ John Dang ⋅ Aditya Grover

Group Preference Optimization: Few-Shot Alignment of Large Language Models

Siyan Zhao ⋅ John Dang ⋅ Aditya Grover

Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression

Junyuan Hong ⋅ Jinhao Duan ⋅ Chenhui Zhang ⋅ Zhangheng LI ⋅ Chulin Xie ⋅ Kelsey Lieberman ⋅ James Diffenderfer ⋅ Brian Bartoldson ⋅ AJAY JAISWAL ⋅ Kaidi Xu ⋅ Bhavya Kailkhura ⋅ Dan Hendrycks ⋅ Dawn Song ⋅ Zhangyang Wang ⋅ Bo Li

Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression

Junyuan Hong ⋅ Jinhao Duan ⋅ Chenhui Zhang ⋅ Zhangheng LI ⋅ Chulin Xie ⋅ Kelsey Lieberman ⋅ James Diffenderfer ⋅ Brian Bartoldson ⋅ AJAY JAISWAL ⋅ Kaidi Xu ⋅ Bhavya Kailkhura ⋅ Dan Hendrycks ⋅ Dawn Song ⋅ Zhangyang Wang ⋅ Bo Li

Leveraging Context in Jailbreaking Attacks

Yixin Cheng ⋅ Markos Georgopoulos ⋅ Volkan Cevher ⋅ Grigorios Chrysos

Leveraging Context in Jailbreaking Attacks

Yixin Cheng ⋅ Markos Georgopoulos ⋅ Volkan Cevher ⋅ Grigorios Chrysos

Self-Alignment of Large Language Models via Social Scene Simulation

Xianghe Pang ⋅ Shuo Tang ⋅ Rui Ye ⋅ Yuxin Xiong ⋅ Bolun Zhang ⋅ Yanfeng Wang ⋅ Siheng Chen

Self-Alignment of Large Language Models via Social Scene Simulation

Xianghe Pang ⋅ Shuo Tang ⋅ Rui Ye ⋅ Yuxin Xiong ⋅ Bolun Zhang ⋅ Yanfeng Wang ⋅ Siheng Chen

Initial Response Selection for Prompt Jailbreaking using Model Steering

Thien Tran ⋅ Koki Wataoka ⋅ Tsubasa Takahashi

Initial Response Selection for Prompt Jailbreaking using Model Steering

Thien Tran ⋅ Koki Wataoka ⋅ Tsubasa Takahashi

Are Large Language Models Bayesian? A Martingale Perspective on In-Context Learning

Fabian Falck ⋅ Ziyu Wang ⋅ Christopher Holmes

Are Large Language Models Bayesian? A Martingale Perspective on In-Context Learning

Fabian Falck ⋅ Ziyu Wang ⋅ Christopher Holmes

Attacks on Third-Party APIs of Large Language Models

Wanru Zhao ⋅ Vidit Khazanchi ⋅ Haodi Xing ⋅ Xuanli He ⋅ Qiongkai Xu ⋅ Nic Lane

Attacks on Third-Party APIs of Large Language Models

Wanru Zhao ⋅ Vidit Khazanchi ⋅ Haodi Xing ⋅ Xuanli He ⋅ Qiongkai Xu ⋅ Nic Lane

How Susceptible are Large Language Models to Ideological Manipulation?

Kai Chen ⋅ Zihao He ⋅ Jun Yan ⋅ Taiwei Shi ⋅ Kristina Lerman

How Susceptible are Large Language Models to Ideological Manipulation?

Kai Chen ⋅ Zihao He ⋅ Jun Yan ⋅ Taiwei Shi ⋅ Kristina Lerman

Enhancing and Evaluating Logical Reasoning Abilities of Large Language Models

Shujie Deng ⋅ Honghua Dong ⋅ Xujie Si

Enhancing and Evaluating Logical Reasoning Abilities of Large Language Models

Shujie Deng ⋅ Honghua Dong ⋅ Xujie Si

Preventing Memorized Completions through White-Box Filtering

Oam Patel ⋅ Rowan Wang

Preventing Memorized Completions through White-Box Filtering

Oam Patel ⋅ Rowan Wang

Safer-Instruct: Aligning Language Models with Automated Preference Data

Taiwei Shi ⋅ Kai Chen ⋅ Jieyu Zhao

Safer-Instruct: Aligning Language Models with Automated Preference Data

Taiwei Shi ⋅ Kai Chen ⋅ Jieyu Zhao

Tailoring Self-Rationalizers with Multi-Reward Distillation

Sahana Ramnath ⋅ Brihi Joshi ⋅ Skyler Hallinan ⋅ Ximing Lu ⋅ Liunian Li ⋅ Aaron Chan ⋅ Jack Hessel ⋅ Yejin Choi ⋅ Xiang Ren

Tailoring Self-Rationalizers with Multi-Reward Distillation

Sahana Ramnath ⋅ Brihi Joshi ⋅ Skyler Hallinan ⋅ Ximing Lu ⋅ Liunian Li ⋅ Aaron Chan ⋅ Jack Hessel ⋅ Yejin Choi ⋅ Xiang Ren

GUARD: Role-playing to Generate Natural-language Jailbreakings to Test Guideline Adherence of Large Language Models

Haibo Jin ⋅ Ruoxi Chen ⋅ Andy Zhou ⋅ Yang Zhang ⋅ Haohan Wang

GUARD: Role-playing to Generate Natural-language Jailbreakings to Test Guideline Adherence of Large Language Models

Haibo Jin ⋅ Ruoxi Chen ⋅ Andy Zhou ⋅ Yang Zhang ⋅ Haohan Wang

Shadowcast: Stealthy Data Poisoning Attacks Against Vision-Language Models

Yuancheng Xu ⋅ Jiarui Yao ⋅ Manli Shu ⋅ Yanchao Sun ⋅ Zichu Wu ⋅ Ning Yu ⋅ Tom Goldstein ⋅ Furong Huang

Shadowcast: Stealthy Data Poisoning Attacks Against Vision-Language Models

Yuancheng Xu ⋅ Jiarui Yao ⋅ Manli Shu ⋅ Yanchao Sun ⋅ Zichu Wu ⋅ Ning Yu ⋅ Tom Goldstein ⋅ Furong Huang

WinoViz: Probing Visual Properties of Objects Under Different States

Woojeong Jin ⋅ Tejas Srinivasan ⋅ Jesse Thomason ⋅ Xiang Ren

WinoViz: Probing Visual Properties of Objects Under Different States

Woojeong Jin ⋅ Tejas Srinivasan ⋅ Jesse Thomason ⋅ Xiang Ren

TrustScore: Reference-Free Evaluation of LLM Response Trustworthiness

Danna Zheng ⋅ Danyang Liu ⋅ Mirella Lapata ⋅ J Pan

TrustScore: Reference-Free Evaluation of LLM Response Trustworthiness

Danna Zheng ⋅ Danyang Liu ⋅ Mirella Lapata ⋅ J Pan

Fight Back Against Jailbreaking via Prompt Adversarial Tuning

Yichuan Mo ⋅ Yuji Wang ⋅ Zeming Wei ⋅ Yisen Wang

Fight Back Against Jailbreaking via Prompt Adversarial Tuning

Yichuan Mo ⋅ Yuji Wang ⋅ Zeming Wei ⋅ Yisen Wang

Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts

Mikayel Samvelyan ⋅ Sharath Raparthy ⋅ Andrei Lupu ⋅ Eric Hambro ⋅ Aram Markosyan ⋅ Manish Bhatt ⋅ Yuning Mao ⋅ Minqi Jiang ⋅ Jack Parker-Holder ⋅ Jakob Foerster ⋅ Tim Rocktaeschel ⋅ Roberta Raileanu

Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts

Mikayel Samvelyan ⋅ Sharath Raparthy ⋅ Andrei Lupu ⋅ Eric Hambro ⋅ Aram Markosyan ⋅ Manish Bhatt ⋅ Yuning Mao ⋅ Minqi Jiang ⋅ Jack Parker-Holder ⋅ Jakob Foerster ⋅ Tim Rocktaeschel ⋅ Roberta Raileanu

A closer look at adversarial suffix learning for Jailbreaking LLMs

Zhe Wang ⋅ Yanjun Qi

A closer look at adversarial suffix learning for Jailbreaking LLMs

Zhe Wang ⋅ Yanjun Qi

Exploring the Adversarial Capabilities of Large Language Models

Lukas Struppek ⋅ Minh Le ⋅ Dominik Hintersdorf ⋅ Kristian Kersting

Exploring the Adversarial Capabilities of Large Language Models

Lukas Struppek ⋅ Minh Le ⋅ Dominik Hintersdorf ⋅ Kristian Kersting

Simple Permutations Can Fool LLaMA: Permutation Attack and Defense for Large Language Models

Liang Chen ⋅ Yatao Bian ⋅ Li Shen ⋅ Kam-Fai Wong

Simple Permutations Can Fool LLaMA: Permutation Attack and Defense for Large Language Models

Liang Chen ⋅ Yatao Bian ⋅ Li Shen ⋅ Kam-Fai Wong

On Prompt-Driven Safeguarding for Large Language Models

Chujie Zheng ⋅ Fan Yin ⋅ Hao Zhou ⋅ Fandong Meng ⋅ Jie Zhou ⋅ Kai-Wei Chang ⋅ Minlie Huang ⋅ Nanyun (Violet) Peng

On Prompt-Driven Safeguarding for Large Language Models

Chujie Zheng ⋅ Fan Yin ⋅ Hao Zhou ⋅ Fandong Meng ⋅ Jie Zhou ⋅ Kai-Wei Chang ⋅ Minlie Huang ⋅ Nanyun (Violet) Peng

Differentially Private Synthetic Data via Foundation Model APIs 2: Text

Chulin Xie ⋅ Zinan Lin ⋅ Arturs Backurs ⋅ Sivakanth Gopi ⋅ Da Yu ⋅ Huseyin Inan ⋅ Harsha Nori ⋅ Haotian Jiang ⋅ Huishuai Zhang ⋅ Yin Tat Lee ⋅ Bo Li ⋅ Sergey Yekhanin

Differentially Private Synthetic Data via Foundation Model APIs 2: Text

Chulin Xie ⋅ Zinan Lin ⋅ Arturs Backurs ⋅ Sivakanth Gopi ⋅ Da Yu ⋅ Huseyin Inan ⋅ Harsha Nori ⋅ Haotian Jiang ⋅ Huishuai Zhang ⋅ Yin Tat Lee ⋅ Bo Li ⋅ Sergey Yekhanin

Watermark Stealing in Large Language Models

Nikola Jovanović ⋅ Robin Staab ⋅ Martin Vechev

Watermark Stealing in Large Language Models

Nikola Jovanović ⋅ Robin Staab ⋅ Martin Vechev

WatME: Towards Lossless Watermarking Through Lexical Redundancy

Liang Chen ⋅ Yatao Bian ⋅ Yang Deng ⋅ Deng Cai ⋅ Shuaiyi Li ⋅ Peilin Zhao ⋅ Kam-Fai Wong

WatME: Towards Lossless Watermarking Through Lexical Redundancy

Liang Chen ⋅ Yatao Bian ⋅ Yang Deng ⋅ Deng Cai ⋅ Shuaiyi Li ⋅ Peilin Zhao ⋅ Kam-Fai Wong

I'm not familiar with the name Harry Potter: Prompting Baselines for Unlearning in LLMs

Pratiksha Thaker ⋅ Yash Maurya ⋅ Virginia Smith

I'm not familiar with the name Harry Potter: Prompting Baselines for Unlearning in LLMs

Pratiksha Thaker ⋅ Yash Maurya ⋅ Virginia Smith

What's in Your "Safe" Data?: Identifying Benign Data that Breaks Safety

Luxi He ⋅ Mengzhou Xia ⋅ Peter Henderson

What's in Your "Safe" Data?: Identifying Benign Data that Breaks Safety

Luxi He ⋅ Mengzhou Xia ⋅ Peter Henderson

Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications

Boyi Wei ⋅ Kaixuan Huang ⋅ Yangsibo Huang ⋅ Tinghao Xie ⋅ Xiangyu Qi ⋅ Mengzhou Xia ⋅ Prateek Mittal ⋅ Mengdi Wang ⋅ Peter Henderson

Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications

Boyi Wei ⋅ Kaixuan Huang ⋅ Yangsibo Huang ⋅ Tinghao Xie ⋅ Xiangyu Qi ⋅ Mengzhou Xia ⋅ Prateek Mittal ⋅ Mengdi Wang ⋅ Peter Henderson

ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs

Fengqing Jiang ⋅ Zhangchen Xu ⋅ Luyao Niu ⋅ Zhen Xiang ⋅ Bhaskar Ramasubramanian ⋅ Bo Li ⋅ Radha Poovendran

ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs

Fengqing Jiang ⋅ Zhangchen Xu ⋅ Luyao Niu ⋅ Zhen Xiang ⋅ Bhaskar Ramasubramanian ⋅ Bo Li ⋅ Radha Poovendran

Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks

Andy Zhou ⋅ Bo Li ⋅ Haohan Wang

Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks

Andy Zhou ⋅ Bo Li ⋅ Haohan Wang

Mechanistically analyzing the effects of fine-tuning on procedurally defined tasks

Samyak Jain ⋅ Robert Kirk ⋅ Ekdeep Singh Lubana ⋅ Robert Dick ⋅ Hidenori Tanaka ⋅ Edward Grefenstette ⋅ Tim Rocktaeschel ⋅ David Krueger

Mechanistically analyzing the effects of fine-tuning on procedurally defined tasks

Samyak Jain ⋅ Robert Kirk ⋅ Ekdeep Singh Lubana ⋅ Robert Dick ⋅ Hidenori Tanaka ⋅ Edward Grefenstette ⋅ Tim Rocktaeschel ⋅ David Krueger

Bayesian reward models for LLM alignment

Adam Yang ⋅ Maxime Robeyns ⋅ Thomas Coste ⋅ Jun Wang ⋅ Haitham Bou Ammar ⋅ Laurence Aitchison

Bayesian reward models for LLM alignment

Adam Yang ⋅ Maxime Robeyns ⋅ Thomas Coste ⋅ Jun Wang ⋅ Haitham Bou Ammar ⋅ Laurence Aitchison

Character-level robustness should be revisited

Elias Abad Rocamora ⋅ Yongtao Wu ⋅ Fanghui Liu ⋅ Grigorios Chrysos ⋅ Volkan Cevher

Character-level robustness should be revisited

Elias Abad Rocamora ⋅ Yongtao Wu ⋅ Fanghui Liu ⋅ Grigorios Chrysos ⋅ Volkan Cevher

Backward Chaining Circuits in a Transformer Trained on a Symbolic Reasoning Task

Jannik Brinkmann ⋅ Abhay Sheshadri ⋅ Victor Levoso ⋅ Paul Swoboda ⋅ Christian Bartelt

Backward Chaining Circuits in a Transformer Trained on a Symbolic Reasoning Task

Jannik Brinkmann ⋅ Abhay Sheshadri ⋅ Victor Levoso ⋅ Paul Swoboda ⋅ Christian Bartelt

Coercing LLMs to do and reveal (almost) anything

Jonas Geiping ⋅ Alex Stein ⋅ Manli Shu ⋅ Khalid Saifullah ⋅ Yuxin Wen ⋅ Tom Goldstein

Coercing LLMs to do and reveal (almost) anything

Jonas Geiping ⋅ Alex Stein ⋅ Manli Shu ⋅ Khalid Saifullah ⋅ Yuxin Wen ⋅ Tom Goldstein

LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B

Simon Lermen ⋅ Charlie Rogers-Smith

LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B

Simon Lermen ⋅ Charlie Rogers-Smith

Assessing Prompt Injection Risks in 200+ Custom GPTs

Jiahao Yu ⋅ Yuhang Wu ⋅ Dong Shu ⋅ Mingyu Jin ⋅ Sabrina Yang ⋅ Xinyu Xing

Assessing Prompt Injection Risks in 200+ Custom GPTs

Jiahao Yu ⋅ Yuhang Wu ⋅ Dong Shu ⋅ Mingyu Jin ⋅ Sabrina Yang ⋅ Xinyu Xing

DUAW: Data-free Universal Adversarial Watermark against Stable Diffusion Customization

Xiaoyu Ye ⋅ Hao Huang ⋅ Jiaqi An ⋅ Yongtao Wang

DUAW: Data-free Universal Adversarial Watermark against Stable Diffusion Customization

Xiaoyu Ye ⋅ Hao Huang ⋅ Jiaqi An ⋅ Yongtao Wang

Red Teaming GPT-4V: Are GPT-4V Safe Against Uni/Multi-Modal Jailbreak Attacks?

Shuo Chen ⋅ Zhen Han ⋅ Bailan He ⋅ Zifeng Ding ⋅ Wenqian Yu ⋅ Philip Torr ⋅ Volker Tresp ⋅ Jindong Gu

Red Teaming GPT-4V: Are GPT-4V Safe Against Uni/Multi-Modal Jailbreak Attacks?

Shuo Chen ⋅ Zhen Han ⋅ Bailan He ⋅ Zifeng Ding ⋅ Wenqian Yu ⋅ Philip Torr ⋅ Volker Tresp ⋅ Jindong Gu

SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding

Zhangchen Xu ⋅ Fengqing Jiang ⋅ Luyao Niu ⋅ Jinyuan Jia ⋅ Bill Yuchen Lin ⋅ Radha Poovendran

SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding

Zhangchen Xu ⋅ Fengqing Jiang ⋅ Luyao Niu ⋅ Jinyuan Jia ⋅ Bill Yuchen Lin ⋅ Radha Poovendran

Retrieval Augmented Prompt Optimization

Yifan Sun ⋅ Jean-Baptiste Tien ⋅ Karthik lakshmanan

Retrieval Augmented Prompt Optimization

Yifan Sun ⋅ Jean-Baptiste Tien ⋅ Karthik lakshmanan

Sequence-Level Certainty Reduces Hallucination In Knowledge-Grounded Dialogue Generation

Yixin Wan ⋅ Fanyou Wu ⋅ Weijie Xu ⋅ Srinivasan Sengamedu

Sequence-Level Certainty Reduces Hallucination In Knowledge-Grounded Dialogue Generation

Yixin Wan ⋅ Fanyou Wu ⋅ Weijie Xu ⋅ Srinivasan Sengamedu

On Trojan Signatures in Large Language Models of Code

Aftab Hussain ⋅ Md Rafiqul Islam Rabin ⋅ Amin Alipour

On Trojan Signatures in Large Language Models of Code

Aftab Hussain ⋅ Md Rafiqul Islam Rabin ⋅ Amin Alipour

Steering LLMs Towards Unbiased Responses: A Causality-Guided Debiasing Framework

Jingling Li ⋅ Zeyu Tang ⋅ Xiaoyu Liu ⋅ Peter Spirtes ⋅ Kun Zhang ⋅ Liu Leqi ⋅ Yang Liu

Steering LLMs Towards Unbiased Responses: A Causality-Guided Debiasing Framework

Jingling Li ⋅ Zeyu Tang ⋅ Xiaoyu Liu ⋅ Peter Spirtes ⋅ Kun Zhang ⋅ Liu Leqi ⋅ Yang Liu

Walk the Talk? Measuring the Faithfulness of Large Language Model Explanations

Katie Matton ⋅ Robert Ness ⋅ Emre Kiciman

Walk the Talk? Measuring the Faithfulness of Large Language Model Explanations

Katie Matton ⋅ Robert Ness ⋅ Emre Kiciman

On Fairness Implications and Evaluations of Low-Rank Adaptation of Large Models

Ken Liu ⋅ Zhoujie Ding ⋅ Berivan Isik ⋅ Sanmi Koyejo

On Fairness Implications and Evaluations of Low-Rank Adaptation of Large Models

Ken Liu ⋅ Zhoujie Ding ⋅ Berivan Isik ⋅ Sanmi Koyejo

CollabEdit: Towards Non-destructive Collaborative Knowledge Editing

Jiamu Zheng ⋅ Jinghuai Zhang ⋅ Futing Wang ⋅ Tianyu Du ⋅ Tao Lin

CollabEdit: Towards Non-destructive Collaborative Knowledge Editing

Jiamu Zheng ⋅ Jinghuai Zhang ⋅ Futing Wang ⋅ Tianyu Du ⋅ Tao Lin

MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs

Yavuz Faruk Bakman ⋅ Duygu Nur Yaldiz ⋅ Baturalp Buyukates ⋅ Chenyang Tao ⋅ Dimitrios Dimitriadis ⋅ Salman Avestimehr

MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs

Yavuz Faruk Bakman ⋅ Duygu Nur Yaldiz ⋅ Baturalp Buyukates ⋅ Chenyang Tao ⋅ Dimitrios Dimitriadis ⋅ Salman Avestimehr

Attacking LLM Watermarks by Exploiting Their Strengths

Qi Pang ⋅ Shengyuan Hu ⋅ Wenting Zheng ⋅ Virginia Smith

Attacking LLM Watermarks by Exploiting Their Strengths

Qi Pang ⋅ Shengyuan Hu ⋅ Wenting Zheng ⋅ Virginia Smith

Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models

Hanlin Zhang ⋅ Benjamin Edelman ⋅ Danilo Francati ⋅ Daniele Venturi ⋅ Giuseppe Ateniese ⋅ Boaz Barak

Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models

Hanlin Zhang ⋅ Benjamin Edelman ⋅ Danilo Francati ⋅ Daniele Venturi ⋅ Giuseppe Ateniese ⋅ Boaz Barak

Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Control

Aleksandar Makelov ⋅ Georg Lange ⋅ Neel Nanda

Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Control

Aleksandar Makelov ⋅ Georg Lange ⋅ Neel Nanda

Privacy-preserving Fine-tuning of Large Language Models through Flatness

Tiejin Chen ⋅ Longchao Da ⋅ Huixue Zhou ⋅ Pingzhi Li ⋅ Kaixiong Zhou ⋅ Tianlong Chen ⋅ Hua Wei

Privacy-preserving Fine-tuning of Large Language Models through Flatness

Tiejin Chen ⋅ Longchao Da ⋅ Huixue Zhou ⋅ Pingzhi Li ⋅ Kaixiong Zhou ⋅ Tianlong Chen ⋅ Hua Wei

Large Language Model Bias Mitigation from the Perspective of Knowledge Editing

Ruizhe Chen ⋅ Yichen Li ⋅ Zikai Xiao ⋅ Zuozhu Liu

Large Language Model Bias Mitigation from the Perspective of Knowledge Editing

Ruizhe Chen ⋅ Yichen Li ⋅ Zikai Xiao ⋅ Zuozhu Liu

Single-pass detection of jailbreaking input in large language models

Leyla Naz Candogan ⋅ Yongtao Wu ⋅ Elias Abad Rocamora ⋅ Grigorios Chrysos ⋅ Volkan Cevher

Single-pass detection of jailbreaking input in large language models

Leyla Naz Candogan ⋅ Yongtao Wu ⋅ Elias Abad Rocamora ⋅ Grigorios Chrysos ⋅ Volkan Cevher

How many Opinions does your LLM have? Improving Uncertainty Estimation in NLG

Lukas Aichberger ⋅ Kajetan Schweighofer ⋅ Mykyta Ielanskyi ⋅ Sepp Hochreiter

How many Opinions does your LLM have? Improving Uncertainty Estimation in NLG

Lukas Aichberger ⋅ Kajetan Schweighofer ⋅ Mykyta Ielanskyi ⋅ Sepp Hochreiter

Self-evaluation and self-prompting to improve the reliability of LLMs

Alexandre Piche ⋅ Aristides Milios ⋅ Dzmitry Bahdanau ⋅ Christopher Pal

Self-evaluation and self-prompting to improve the reliability of LLMs

Alexandre Piche ⋅ Aristides Milios ⋅ Dzmitry Bahdanau ⋅ Christopher Pal

BEYOND FINE-TUNING: LORA MODULES BOOST NEAR- OOD DETECTION AND LLM SECURITY

Etienne Salimbeni ⋅ Francesco Craighero ⋅ Renata Khasanova ⋅ Milos Vasic ⋅ Pierre Vandergheynst

BEYOND FINE-TUNING: LORA MODULES BOOST NEAR- OOD DETECTION AND LLM SECURITY

Etienne Salimbeni ⋅ Francesco Craighero ⋅ Renata Khasanova ⋅ Milos Vasic ⋅ Pierre Vandergheynst

PETA: PARAMETER-EFFICIENT TROJAN ATTACKS

Lauren Hong ⋅ Ting Wang

PETA: PARAMETER-EFFICIENT TROJAN ATTACKS

Lauren Hong ⋅ Ting Wang

Is Your Jailbreaking Prompt Truly Effective for Large Language Models?

Bochuan Cao ⋅ Tianrong Zhang ⋅ Yuanpu Cao ⋅ Jinyuan Jia ⋅ Lu Lin ⋅ Jinghui Chen

Is Your Jailbreaking Prompt Truly Effective for Large Language Models?

Bochuan Cao ⋅ Tianrong Zhang ⋅ Yuanpu Cao ⋅ Jinyuan Jia ⋅ Lu Lin ⋅ Jinghui Chen

TOFU: A Task of Fictitious Unlearning for LLMs

Pratyush Maini ⋅ Zhili Feng ⋅ Avi Schwarzschild ⋅ Zachary Lipton ⋅ J Kolter

TOFU: A Task of Fictitious Unlearning for LLMs

Pratyush Maini ⋅ Zhili Feng ⋅ Avi Schwarzschild ⋅ Zachary Lipton ⋅ J Kolter

The Effect of Model Size on LLM Post-hoc Explainability via LIME

Henning Heyen ⋅ Amy Widdicombe ⋅ Noah Siegel ⋅ Philip Treleaven ⋅ Maria Perez-Ortiz

The Effect of Model Size on LLM Post-hoc Explainability via LIME

Henning Heyen ⋅ Amy Widdicombe ⋅ Noah Siegel ⋅ Philip Treleaven ⋅ Maria Perez-Ortiz

Calibrating Language Models With Adaptive Temperature Scaling

Johnathan Xie ⋅ Annie Chen ⋅ Yoonho Lee ⋅ Eric Mitchell ⋅ Chelsea Finn

Calibrating Language Models With Adaptive Temperature Scaling

Johnathan Xie ⋅ Annie Chen ⋅ Yoonho Lee ⋅ Eric Mitchell ⋅ Chelsea Finn

Source-Aware Training Enables Knowledge Attribution in Language Models

Muhammad Khalifa ⋅ David Wadden ⋅ Emma Strubell ⋅ Honglak Lee ⋅ Lu Wang ⋅ Iz Beltagy ⋅ Hao Peng

Source-Aware Training Enables Knowledge Attribution in Language Models

Muhammad Khalifa ⋅ David Wadden ⋅ Emma Strubell ⋅ Honglak Lee ⋅ Lu Wang ⋅ Iz Beltagy ⋅ Hao Peng

Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models

Xianjun Yang ⋅ Xiao Wang ⋅ Qi Zhang ⋅ Linda Petzold ⋅ William Wang ⋅ XUN ZHAO ⋅ Dahua Lin

Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models

Xianjun Yang ⋅ Xiao Wang ⋅ Qi Zhang ⋅ Linda Petzold ⋅ William Wang ⋅ XUN ZHAO ⋅ Dahua Lin

PANDORA: Detailed LLM Jailbreaking via Collaborated Phishing Agents with Decomposed Reasoning

Zhaorun Chen ⋅ Zhuokai Zhao ⋅ Wenjie Qu ⋅ zichen wen ⋅ Zhiguang Han ⋅ Zhihong Zhu ⋅ Jiaheng Zhang ⋅ Huaxiu Yao

PANDORA: Detailed LLM Jailbreaking via Collaborated Phishing Agents with Decomposed Reasoning

Zhaorun Chen ⋅ Zhuokai Zhao ⋅ Wenjie Qu ⋅ zichen wen ⋅ Zhiguang Han ⋅ Zhihong Zhu ⋅ Jiaheng Zhang ⋅ Huaxiu Yao

Break it, Imitate it, Fix it: Robustness by Generating Human-Like Attacks

Aradhana Sinha ⋅ Ananth Balashankar ⋅ Ahmad Beirami ⋅ Thi Avrahami ⋅ Jilin Chen ⋅ Alex Beutel

Break it, Imitate it, Fix it: Robustness by Generating Human-Like Attacks

Aradhana Sinha ⋅ Ananth Balashankar ⋅ Ahmad Beirami ⋅ Thi Avrahami ⋅ Jilin Chen ⋅ Alex Beutel

Quantitative Certification of Knowledge Comprehension in LLMs

Isha Chaudhary ⋅ Vedaant Jain ⋅ Gagandeep Singh

Quantitative Certification of Knowledge Comprehension in LLMs

Isha Chaudhary ⋅ Vedaant Jain ⋅ Gagandeep Singh

Toward Robust Unlearning for LLMs

Rishub Tamirisa ⋅ Bhrugu Bharathi ⋅ Andy Zhou ⋅ Bo Li ⋅ Mantas Mazeika

Toward Robust Unlearning for LLMs

Rishub Tamirisa ⋅ Bhrugu Bharathi ⋅ Andy Zhou ⋅ Bo Li ⋅ Mantas Mazeika

Explorations of Self-Repair in Language Model

Cody Rushing ⋅ Neel Nanda

Explorations of Self-Repair in Language Model

Cody Rushing ⋅ Neel Nanda

Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?

Egor Zverev ⋅ Sahar Abdelnabi ⋅ Mario Fritz ⋅ Christoph Lampert

Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?

Egor Zverev ⋅ Sahar Abdelnabi ⋅ Mario Fritz ⋅ Christoph Lampert

An Assessment of Model-on-Model Deception

Julius Heitkoetter ⋅ Michael Gerovitch ⋅ Laker Newhouse

An Assessment of Model-on-Model Deception

Julius Heitkoetter ⋅ Michael Gerovitch ⋅ Laker Newhouse

Open Sesame! Universal Black-Box Jailbreaking of Large Language Models

Raz Lapid ⋅ Ron Langberg ⋅ Moshe Sipper

Open Sesame! Universal Black-Box Jailbreaking of Large Language Models

Raz Lapid ⋅ Ron Langberg ⋅ Moshe Sipper