← Search

Zhiqing Sun

24 accepted papers

2026

Toward Training Superintelligent Software Agents through Self-Play SWE-RL

ICML 2026poster

While current software agents powered by large language models (LLMs) and reinforcement learning (RL) can boost programmer productivity, their reliance on human-curated training data and environments creates a fundamental barrier to superintelligence. In this paper, we present Self-play SWE-RL (SSR)…

Cited by 0SourceScholar
2025

Direct Preference Optimization of Video Large Multimodal Models from Language Model Reward

NAACL 2025long

Preference modeling techniques, such as direct preference optimization (DPO), has shown effective in enhancing the generalization abilities of large language model (LLM). However, in tasks involving video instruction-following, providing informative feedback, especially for open-ended conversations,…

2025

Improve Vision Language Model Chain-of-thought Reasoning

ACL 2025long

Chain-of-thought (CoT) reasoning in vision language models (VLMs) is crucial for improving interpretability and trustworthiness. However, current training recipes often relying on datasets dominated by short annotations with minimal rationales. In this work, we show that training VLM on short answer…

2025

Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for LLM Problem-Solving

ICLR 2025poster

While the scaling laws of large language models (LLMs) training have been extensively studied, optimal inference configurations of LLMs remain underexplored. We study inference scaling laws (aka test-time scaling laws) and compute-optimal inference, focusing on the trade-offs between model sizes an…

Cited by 10SourcePDFScholar
2025

Self-Play Preference Optimization for Language Model Alignment

ICLR 2025poster

Standard reinforcement learning from human feedback (RLHF) approaches relying on parametric models like the Bradley-Terry model fall short in capturing the intransitivity and irrationality in human preferences. Recent advancements suggest that directly working with preference probabilities can yield…

2024

Aligning Large Multimodal Models with Factually Augmented RLHF

ACL 2024findings

Large Multimodal Models (LMM) are built across modalities and the misalignment between two modalities can result in “hallucination”, generating textual outputs that are not grounded by the multimodal information in context. To address the multimodal misalignment issue, we adapt the Reinforcement Lea…

2024

Easy-to-Hard Generalization: Scalable Alignment Beyond Human Supervision

NeurIPS 2024poster

Current AI alignment methodologies rely on human-provided demonstrations or judgments, and the learned capabilities of AI systems would be upper-bounded by human capabilities as a result. This raises a challenging research question: How can we keep improving the systems when their capabilities have…

2024

Instruction-tuned Language Models are Better Knowledge Learners

ACL 2024long

In order for large language model (LLM)-based assistants to effectively adapt to evolving information needs, it must be possible to update their factual knowledge through continued training on new data. The standard recipe for doing so involves continued pre-training on new documents followed by ins…

2024

LLM and Simulation as Bilevel Optimizers: A New Paradigm to Advance Physical Scientific Discovery

ICML 2024poster

Large Language Models have recently gained significant attention in scientific discovery for their extensive knowledge and advanced reasoning capabilities. However, they encounter challenges in effectively simulating observational feedback and grounding it with language to propel advancements in phy…

2024

SALMON: Self-Alignment with Instructable Reward Models

ICLR 2024poster

Supervised Fine-Tuning (SFT) on response demonstrations combined with Reinforcement Learning from Human Feedback (RLHF) constitutes a powerful paradigm for aligning LLM-based AI agents. However, a significant limitation of such an approach is its dependency on high-quality human annotations, making…

2024

Visual Chain-of-Thought Prompting for Knowledge-Based Visual Reasoning

AAAI 2024technical

Knowledge-based visual reasoning remains a daunting task since it not only requires machines to interpret the concepts and relationships from visual scenes but also associate them with external world knowledge to conduct a chain of reasoning on open-world questions. Previous works, however, treat vi…

2023

Active Retrieval Augmented Generation

EMNLP 2023long main

Despite the remarkable ability of large language models (LMs) to comprehend and generate language, they have a tendency to hallucinate and create factually inaccurate output. Augmenting LMs by retrieving information from external knowledge resources is one promising solution. Most existing retrieval…

Cited by 0SourcecodeScholar
2023

Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervision

NeurIPS 2023spotlight

Recent AI-assistant agents, such as ChatGPT, predominantly rely on supervised fine-tuning (SFT) with human annotations and reinforcement learning from human feedback (RLHF) to align the output of large language models (LLMs) with human intentions, ensuring they are helpful, ethical, and reliable. Ho…

2022

DIMES: A Differentiable Meta Solver for Combinatorial Optimization Problems

NeurIPS 2022accept

Recently, deep reinforcement learning (DRL) models have shown promising results in solving NP-hard Combinatorial Optimization (CO) problems. However, most DRL solvers can only scale to a few hundreds of nodes for combinatorial optimization problems on graphs, such as the Traveling Salesman Problem (…

2021

Rethinking Transformer-Based Set Prediction for Object Detection

ICCV 2021poster

DETR is a recently proposed Transformer-based method which views object detection as a set prediction problem and achieves state-of-the-art performance but demands extra-long training time to converge. In this paper, we investigate the causes of the optimization difficulty in the training of DETR. O…

Cited by 425PDFcodeScholar
2020

Dynamically Pruned Message Passing Networks for Large-scale Knowledge Graph Reasoning

ICLR 2020poster

We propose Dynamically Pruned Message Passing Networks (DPMPN) for large-scale knowledge graph reasoning. In contrast to existing models, embedding-based or path-based, we learn an input-dependent subgraph to explicitly model a sequential reasoning process. Each subgraph is dynamically constructed,…

Cited by 86SourcecodeScholar
2019

Fast Structured Decoding for Sequence Models

NeurIPS 2019poster

Autoregressive sequence models achieve state-of-the-art performance in domains like machine translation. However, due to the autoregressive factorization nature, these models suffer from heavy latency during inference. Recently, non-autoregressive sequence models were proposed to speed up the infere…

2019

RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space

ICLR 2019poster

We study the problem of learning representations of entities and relations in knowledge graphs for predicting missing links. The success of such a task heavily relies on the ability of modeling and inferring the patterns of (or between) the relations. In this paper, we present a new approach for kno…