ICLR

Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning

Mingyang Chen, Haoze Sun, Tianpeng Li, Fan Yang, Hao Liang, Keer Lu, Bin Cui, Wentao Zhang, Zenan Zhou, Weipeng Chen

Comments Accepted to ICLR 2025

Data-adaptive Differentially Private Prompt Synthesis for In-Context Learning

Fengyu Gao, Ruida Zhou, Tianhao Wang, Cong Shen, Jing Yang

Comments Accepted to ICLR 2025

Improving Long-Text Alignment for Text-to-Image Diffusion Models

Luping Liu, Chao Du, Tianyu Pang, Zehan Wang, Chongxuan Li, Dong Xu

Journal ref International Conference on Learning Representations (ICLR 2025)

When Attention Sink Emerges in Language Models: An Empirical View

Xiangming Gu, Tianyu Pang, Chao Du, Qian Liu, Fengzhuo Zhang, Cunxiao Du, Ye Wang, Min Lin

Comments ICLR 2025 (Spotlight)

CtrLoRA: An Extensible and Efficient Framework for Controllable Image Generation

Yifeng Xu, Zhenliang He, Shiguang Shan, Xilin Chen

Comments ICLR 2025. Code: https://github.com/xyfJASON/ctrlora

Poison-splat: Computation Cost Attack on 3D Gaussian Splatting

Jiahao Lu, Yifan Zhang, Qiuhong Shen, Xinchao Wang, Shuicheng Yan

Comments Accepted by ICLR 2025 as a spotlight paper

MACPO: Weak-to-Strong Alignment via Multi-Agent Contrastive Preference Optimization

Yougang Lyu, Lingyong Yan, Zihan Wang, Dawei Yin, Pengjie Ren, Maarten de Rijke, Zhaochun Ren

Comments ICLR 2025

IterGen: Iterative Semantic-aware Structured LLM Generation with Backtracking

Shubham Ugare, Rohan Gumaste, Tarun Suresh, Gagandeep Singh, Sasa Misailovic

Comments Accepted at ICLR 2025

Sylber: Syllabic Embedding Representation of Speech from Raw Audio

Cheol Jun Cho, Nicholas Lee, Akshat Gupta, Dhruv Agarwal, Ethan Chen, Alan W Black, Gopala K. Anumanchipalli

Comments Accepted at ICLR 2025

Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates

Xiaosen Zheng, Tianyu Pang, Chao Du, Qian Liu, Jing Jiang, Min Lin

Comments ICLR 2025 (Oral)

SymDiff: Equivariant Diffusion via Stochastic Symmetrisation

Leo Zhang, Kianoosh Ashouritaklimi, Yee Whye Teh, Rob Cornish

Comments Camera-ready version for ICLR 2025

From Tokens to Words: On the Inner Lexicon of LLMs

Guy Kaplan, Matanel Oren, Yuval Reif, Roy Schwartz

Comments Accepted to the International Conference on Learning Representations (ICLR) 2025

TRACE: Temporal Grounding Video LLM via Causal Event Modeling

Yongxin Guo, Jingyu Liu, Mingda Li, Qingbin Liu, Xi Chen, Xiaoying Tang

Comments ICLR 2025

On the Optimization and Generalization of Two-layer Transformers with Sign Gradient Descent

Bingrui Li, Wei Huang, Andi Han, Zhanpeng Zhou, Taiji Suzuki, Jun Zhu, Jianfei Chen

Comments 79 pages, 19 figures, ICLR 2025 Spotlight

The Optimization Landscape of SGD Across the Feature Learning Strength

Alexander Atanasov, Alexandru Meterez, James B. Simon, Cengiz Pehlevan

Comments ICLR 2025 Final Copy, 40 Pages, 45 figures

Inference Scaling for Long-Context Retrieval Augmented Generation

Zhenrui Yue, Honglei Zhuang, Aijun Bai, Kai Hui, Rolf Jagerman, Hansi Zeng, Zhen Qin, Dong Wang, Xuanhui Wang, Michael Bendersky

Comments ICLR 2025

How Discrete and Continuous Diffusion Meet: Comprehensive Analysis of Discrete Diffusion Models via a Stochastic Integral Framework

Yinuo Ren, Haoxuan Chen, Grant M. Rotskoff, Lexing Ying

Comments Published at ICLR 2025

Steering Large Language Models between Code Execution and Textual Reasoning

Yongchao Chen, Harsh Jhamtani, Srinagesh Sharma, Chuchu Fan, Chi Wang

Comments 32 pages, 12 figures, 12 tables

Journal ref The Thirteenth International Conference on Learning Representations (ICLR'2025)

LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding

Doohyuk Jang, Sihwan Park, June Yong Yang, Yeonsung Jung, Jihun Yun, Souvik Kundu, Sung-Yub Kim, Eunho Yang

Comments 30 pages, 13 figures, Accepted to ICLR 2025 (poster)

X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale

Haoran Xu, Kenton Murray, Philipp Koehn, Hieu Hoang, Akiko Eriguchi, Huda Khayrallah

Comments Published as a conference paper at ICLR 2025 (spotlight)

Towards Understanding the Universality of Transformers for Next-Token Prediction

Michael E. Sander, Gabriel Peyré

Comments ICLR 2025, 20 pages

MANTRA: The Manifold Triangulations Assemblage

Rubén Ballester, Ernst Röell, Daniel Bīn Schmid, Mathieu Alain, Sergio Escalera, Carles Casacuberta, Bastian Rieck

Comments Accepted at ICLR 2025 (https://openreview.net/forum?id=X6y5CC44HM)

Boosting Perturbed Gradient Ascent for Last-Iterate Convergence in Games

Kenshi Abe, Mitsuki Sakamoto, Kaito Ariu, Atsushi Iwasaki

Comments Accepted at ICLR 2025

Structural-Entropy-Based Sample Selection for Efficient and Effective Learning

Tianchi Xie, Jiangning Zhu, Guozu Ma, Minzhi Lin, Wei Chen, Weikai Yang, Shixia Liu

Comments Published as a conference paper at ICLR 2025

Robust Weight Initialization for Tanh Neural Networks with Fixed Point Analysis

Hyunwoo Lee, Hayoung Choi, Hyunju Kim

Comments ICLR 2025

The Labyrinth of Links: Navigating the Associative Maze of Multi-modal LLMs

Hong Li, Nanxi Li, Yuanjie Chen, Jianbin Zhu, Qinlu Guo, Cewu Lu, Yong-Lu Li

Comments Accepted by ICLR 2025. Project page: https://mvig-rhos.com/llm_inception

Score Forgetting Distillation: A Swift, Data-Free Method for Machine Unlearning in Diffusion Models

Tianqi Chen, Shujian Zhang, Mingyuan Zhou

Comments ICLR 2025

HiLo: A Learning Framework for Generalized Category Discovery Robust to Domain Shifts

Hongjun Wang, Sagar Vaze, Kai Han

Comments v2: Accepted as a conference paper at ICLR 2025; Project page: https://github.com/Visual-AI/hilo/

Exploring the Effectiveness of Object-Centric Representations in Visual Question Answering: Comparative Insights with Foundation Models

Amir Mohammad Karimi Mamaghan, Samuele Papa, Karl Henrik Johansson, Stefan Bauer, Andrea Dittadi

Comments Published at ICLR 2025

Generalization v.s. Memorization: Tracing Language Models' Capabilities Back to Pretraining Data

Xinyi Wang, Antonis Antoniades, Yanai Elazar, Alfonso Amayuelas, Alon Albalak, Kexun Zhang, William Yang Wang

Comments Accepted to ICLR 2025