Yijie Jin

Chenkai Xu, Yijie Jin, Jiajun Li et al.

Diffusion Large Language Models (dLLMs) have demonstrated significant potential for high-speed inference. However, current confidence-driven decoding strategies are constrained by limited parallelism, typically achieving only 1--3 tokens per forward pass (TPF). In this work, we identify that the degree of parallelism during dLLM inference is highly sensitive to the Token Filling Order (TFO). Then, we introduce Lookahead PArallel Decoding LoPA, a training-free, plug-and-play algorithm, to identify a superior TFO and hence accelerate inference. LoPA concurrently explores distinct candidate TFOs via parallel branches, and selects the one with the highest potential for future parallelism based on branch confidence. We apply LoPA to the state-of-the-art D2F model and observe a substantial enhancement in decoding efficiency. Notably, LoPA increases the TPF of D2F-Dream to 10.1 on the GSM8K while maintaining performance superior to the Dream baseline. Furthermore, to facilitate this unprecedented degree of parallelism, we develop a specialized multi-device inference system featuring Branch Parallelism (BP), which achieves a single-sample throughput of 1073.9 tokens per second under multi-GPU deployment. The code is available at https://github.com/zhijie-group/LoPA.

Yanzhe Hu, Yijie Jin, Pengfei Liu et al.

Diffusion Large Language Models (dLLMs) have emerged as a promising paradigm for parallel token generation, with block-wise variants garnering significant research interest. Despite their potential, existing dLLMs typically suffer from a rigid accuracy-parallelism trade-off: increasing the number of tokens per forward (TPF) via aggressive parallel decoding often leads to performance degradation and increased generation instability. We identify that this limitation stems from the model's inability to navigate high-parallelism regimes where approximation errors and local corruptions accumulate, ultimately undermining the reliability of parallel generation. To address this, we propose LightningRL, a post-training framework designed to directly optimize the speed-quality Pareto frontier of pre-trained dLLMs. Instead of forcing uniform parallelization, our approach leverages reinforcement learning to identify and reinforce high-parallelism trajectories that maintain generation accuracy. Built upon the Group Relative Policy Optimization (GRPO) framework, LightningRL introduces several enhancements tailored for dLLMs: (1) stabilized training via per-reward decoupled normalization; (2) token-level negative log-likelihood (NLL) regularization on correct trajectories to anchor model performance; and (3) a dynamic sampling strategy with TPF-aware filtering to enhance training efficiency. Experimental results across mathematical and coding benchmarks demonstrate that LightningRL consistently advances the Pareto frontier, achieving competitive task accuracy while significantly increasing parallelism, reaching an average TPF of 7.32 (with a peak of 11.10 on the MBPP dataset). Our code is available at https://github.com/SJTU-DENG-Lab/LightningRL.

Xu Wang, Chenkai Xu, Yijie Jin et al.

Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to autoregressive (AR) LLMs for text generation, with the potential to decode multiple tokens in a single iteration. However, none of the existing open-source dLLMs have achieved superior inference speed over AR LLMs of similar size. This paper breaks this barrier based on a simple and effective strategy named discrete diffusion forcing (D2F). D2F equips dLLMs with two key capabilities: (1) block-wise autoregressive generation to enable KV cache utilization; (2) prediction of following tokens without requiring completion of prior blocks for inter-block parallel decoding. In this way, the vanilla dLLMs are refurbished into an AR-diffusion hybrid paradigm for efficient inference. D2F can be implemented with an asymmetric distillation process based on pre-trained dLLMs. We further propose a pipelined parallel decoding algorithm, which enables a trade-off between efficiency and efficacy. Empirically, D2F dLLMs achieve more than $\mathbf{2.5\times}$ inference speed than LLaMA3 and Qwen2.5 on GSM8K. Compared to vanilla dLLMs like LLaDA and Dream, the acceleration can be more than $\mathbf{50\times}$ while maintaining comparable output quality. The code is available at https://github.com/zhijie-group/Discrete-Diffusion-Forcing.

Chichen Lin, Yijie Jin, Kangbo Hu et al.

Misinformation resilience is a dynamic community process: communities differ not only in whether they initially trust false claims, but also in how they recover through interaction, questioning, correction, and support withdrawal. We study this process with an LLM-based agent simulation that constructs synthetic communities along two theoretically motivated dimensions: Actively Open-minded Thinking (AOT), which captures evidence-seeking and willingness to revise beliefs, and Political Ideology (PI), which captures identity-based interpretation of contested claims. These two traits allow us to examine how evidence-oriented reasoning and ideological alignment jointly shape community responses to credible misinformation shocks. Across systematically varied AOT-PI communities, we find that higher AOT improves both resistance to misinformation uptake and recovery after trust peaks. PI shapes the recovery pathway: ideologically moderate communities recover more reliably, while polarized communities retain more residual support. Stance-level analysis shows that resilience depends on whether agents move from questioning a claim to denying or correcting it and withdrawing prior support. Intervention experiments further show that persuasion and fact checking better support post-peak correction, whereas accuracy prompts mainly induce early caution and source warnings have weaker effects. Together, this work provides a mechanism-level account of community misinformation resilience, showing how psychological composition and intervention design shape whether communities move from misinformation exposure toward correction or persistent support.

Yijie Jin

Multimodal Sentiment Analysis (MSA) leverages multiple data modals to analyze human sentiment. Existing MSA models generally employ cutting-edge multimodal fusion and representation learning-based methods to promote MSA capability. However, there are two key challenges: (i) in existing multimodal fusion methods, the decoupling of modal combinations and tremendous parameter redundancy, lead to insufficient fusion performance and efficiency; (ii) a challenging trade-off exists between representation capability and computational overhead in unimodal feature extractors and encoders. Our proposed GSIFN incorporates two main components to solve these problems: (i) a graph-structured and interlaced-masked multimodal Transformer. It adopts the Interlaced Mask mechanism to construct robust multimodal graph embedding, achieve all-modal-in-one Transformer-based fusion, and greatly reduce the computational overhead; (ii) a self-supervised learning framework with low computational overhead and high performance, which utilizes a parallelized LSTM with matrix memory to enhance non-verbal modal features for unimodal label generation. Evaluated on the MSA datasets CMU-MOSI, CMU-MOSEI, and CH-SIMS, GSIFN demonstrates superior performance with significantly lower computational overhead compared with previous state-of-the-art models.

Yijie Jin, Junjie Peng, Xuanchao Lin et al.

Multimodal Sentiment Analysis (MSA) is a rapidly developing field that integrates multimodal information to recognize sentiments, and existing models have made significant progress in this area. The central challenge in MSA is multimodal fusion, which is predominantly addressed by Multimodal Transformers (MulTs). Although act as the paradigm, MulTs suffer from efficiency concerns. In this work, from the perspective of efficiency optimization, we propose and prove that MulTs are hierarchical modal-wise heterogeneous graphs (HMHGs), and we introduce the graph-structured representation pattern of MulTs. Based on this pattern, we propose an Interlaced Mask (IM) mechanism to design the Graph-Structured and Interlaced-Masked Multimodal Transformer (GsiT). It is formally equivalent to MulTs which achieves an efficient weight-sharing mechanism without information disorder through IM, enabling All-Modal-In-One fusion with only 1/3 of the parameters of pure MulTs. A Triton kernel called Decomposition is implemented to ensure avoiding additional computational overhead. Moreover, it achieves significantly higher performance than traditional MulTs. To further validate the effectiveness of GsiT itself and the HMHG concept, we integrate them into multiple state-of-the-art models and demonstrate notable performance improvements and parameter reduction on widely used MSA datasets.