Papers by Minjia Zhang

8 papers
MiniKV: Pushing the Limits of 2-Bit KV Cache via Compression and System Co-Design for Efficient Long Context Inference (2025.findings-acl)

Copied to clipboard

Challenge: State-of-the-art 2-bit KV cache quantization methods achieve excellent results in accelerating LLM inference while retaining accuracy on long context tasks.
Approach: They propose a method based on 2-bit KV cache quantization with adaptive KV policies that retain LLM accuracy with only a subset of KV states.
Outcome: The proposed method outperforms state-of-the-art methods on a wide range of long context tasks while retaining accuracy.
A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks (2026.acl-long)

Copied to clipboard

Challenge: Recent advances in large language models (LLMs) have leapt from static chatbots to versatile agents that tackle complex tasks such as science experiments.
Approach: They propose a plan-and-execute framework and propose 'EAGLET' to enhance the executor agent's planning abilities without human effort.
Outcome: The proposed method outperforms existing methods on three long-horizon tasks and reduces training costs by 8 compared to baselines.
FaithLens: Detecting and Explaining Faithfulness Hallucination (2026.findings-acl)

Copied to clipboard

Challenge: Recent progress in large language models (LLMs) has revolutionized text generation.
Approach: They propose a faithfulness hallucination detection model that can provide binary predictions and corresponding explanations to improve trustworthiness.
Outcome: The proposed model outperforms advanced models on 12 diverse tasks.
Hidden States as Early Signals: Step-level Trace Evaluation and Pruning for Efficient Test-Time Scaling (2026.findings-acl)

Copied to clipboard

Challenge: Existing approaches to speed up parallel scaling have relied on similarity-based or confidence-based pruning, but these signals do not reliably indicate trace quality.
Approach: They propose a pruning framework that evaluates reasoning steps using hidden states and dynamically prunes unpromising traces during generation.
Outcome: The proposed framework reduces end-to-end inference latency by 45%–70% on average compared to self-consistency while improving reasoning accuracy.
Looking Beyond Text: Reducing Language Bias in Large Vision-Language Models via Multimodal Dual-Attention and Soft-Image Guidance (2025.emnlp-main)

Copied to clipboard

Challenge: Large vision-language models (LVLMs) have been criticized for their language bias.
Approach: They propose to use a dual-attention mechanism to construct separate attention for visual and text inputs to enhance integration of visual inputs across models.
Outcome: Experiments show that the proposed model debiases LVLMs from their language bias, enhancing visual comprehension and reducing hallucinations without additional resources.
MENTOR: Efficient Autoregressive Image Generation with Balanced Multimodal Control (2026.findings-acl)

Copied to clipboard

Challenge: Recent text-to-image models achieve impressive visual quality but still face challenges in precise controllability, balancing multimodal inputs, and high training cost for multimodal image generation.
Approach: They propose an autoregressive framework with a two-stage training paradigm for controllable multimodal image generation.
Outcome: Extensive experiments on DreamBench++ and DreamBech show that the proposed framework achieves a strong balance between textual and visual guidance for controllable image generation.
MedCite: Can Language Models Generate Verifiable Text for Medicine? (2025.findings-acl)

Copied to clipboard

Challenge: Existing LLM-based medical question answering systems lack citation generation and evaluation capabilities, raising concerns about their adoption in practice.
Approach: They propose a framework that facilitates the design and evaluation of LLM citations for medical tasks and a retrieval-citation method that generates high-quality citation.
Outcome: The proposed method achieves superior citation precision and recall improvements compared to strong baseline methods and correlates well with annotation results from professional experts.
Cache-of-Thought: Master-Apprentice Framework for Cost-Effective Vision Language Model Reasoning (2025.emnlp-main)

Copied to clipboard

Challenge: Recent Vision Language Models (VLMs) have shown tremendous promise in a wide range of realworld applications, but their size has made at-scale deployment and operation challenging due to high consumption of cloud computing resource, high latency, and expensive API calls.
Approach: They propose a master–apprentice framework for collaborative inference between large and small vision language models.
Outcome: The proposed framework improves reasoning performance on widely-recognized and challenging general reasoning benchmarks and specifically boosts reasoning of apprentice VLMs by 36.6%.

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