20 papers
ValuSpec: Plug-and-Play Candidate Valuation before Target Verification for Tree-Based Speculative Decoding
Liang He, SiYuan Ma, Qishi Zhan +5
Tree-based speculative decoding accelerates large language model inference by verifying multiple candidate branches in parallel. However, in practical serving scenarios, similar re…
Why On-Policy Distillation Sometimes Fails: Vanishing Learning Signals
Lei Zhao, Qichao Zhao, Bowen Zuo +1
On-policy distillation (OPD) enables effective capability transfer between language models, yet the mechanisms underlying its failures are not fully understood. Across code generat…
When Verifiable Counts Depend on Wording: Auditing Wording Robustness in Instruction Following
Qishi Zhan, Seoyeon Jang, Zihan Dong +3
Verifiable instruction-following benchmarks often express each constraint through one fixed template. We test whether scores remain stable when the operational requirement is uncha…
Behavior Pack Optimization for Video MLLM Post-Training
Zhaolu Kang, Shiyu Liu, Tailong Luo +12
Video multimodal large language models (MLLMs) keep climbing video question answering benchmarks, yet shuffling the frames, masking the segment that supports the answer, or occludi…
Probability is Not Enough: Exploring and Counting Divergent Tokens for Reasoning Uncertainty Quantification in LLMs
Feiyang Li, Shengjing Liu, Qi Zhan +7
As the chain-of-thought reasoning capabilities of large language models improve, evaluating and calibrating their reasoning confidence is becoming increasingly important for quanti…
CADWorld: Computer-Use Benchmark for Long-Horizon Computer-Aided Design
Zihan Dong, Yuanzhe Liu, Zhiyuan Ma +4
Computer-use agents are increasingly evaluated in realistic desktop environments, but existing benchmarks provide limited coverage of professional engineering workflows whose outpu…
Why Are GUI Agents Correct but Late? Decode on the Decision-Time Critical Path, Tested with Pre-Compiled Policy Trees
Zihan Dong, Rui Qian, Qishi Zhan +3
Computer-use agents often fail on transient GUI events because they produce the correct action only after the relevant window has already closed. We identify the main cause as expe…
How Benchmarks Mis-Score Computer-Use Agents
Zihan Dong, Zhiyuan Ma, Zekun Wang +5
Computer-use agents (CUA) are being deployed to browse the web and operate desktop software, yet their benchmark scores are still commonly produced by brittle scripted oracles. A s…
Discontinuous Prior-Mode Sections and the Geometry of Ambiguity in Intrinsic Image Decomposition
Ziheng Chen, Liangchen Liu, Qishi Zhan +2
The viral 2015 photograph known as "The Dress" divides observers into two camps because it is ambiguous: the same image colors can be explained either as a blue-black surface under…
Stress Amplified Resilience: ESG and Joint Fragility in Equity Markets
Minxuan Hu, Jiayu Yi, Ziheng Chen +2
Market stress rarely harms investors through one channel alone. Losses, volatility spikes, and deteriorating tradability often arrive together. We examine whether ESG is associated…
BudgetDraft: Acceptance-Aware Multi-View Training for Sparse-KV Speculative Decoding
Liang He, Jingbo Wen, Qishi Zhan +4
Speculative decoding speeds up autoregressive decoding by using a drafter to propose multiple tokens that a verifier validates in parallel. In resource-constrained deployments, the…
Disentangling Language Roles in Multilingual LLM Task Execution
Qishi Zhan, Minxuan Hu, Seoyeon Jang +7
Multilingual LLMs are increasingly used when instruction, source content, and required response languages do not coincide. Existing benchmarks have expanded multilingual instructio…
Relative Repairability: A Calibration-Based Diagnostic for High-Sparsity Post-Pruning Allocation
Qishi Zhan, Liang He, Minxuan Hu +1
At very high sparsity, neural network pruning does more than decide which weights remain. It also determines where pruning induced damage is placed across the network, and whether…
How Sparsity Allocation Shapes Label-Free Post-Pruning Recoverability
Qishi Zhan, Minxuan Hu, Liang He
Unstructured magnitude pruning at high sparsity can reduce neural network accuracy to near-random performance, while labeled retraining may be unavailable in practical deployment s…
Adaptive Signal Resuscitation: Channel-wise Post-Pruning Repair for Sparse Vision Networks
Qishi Zhan, Ziheng Chen, Minxuan Hu
One-shot magnitude pruning can cause severe accuracy collapse in the high-sparsity regime, even when the pruning mask preserves the largest weights. We argue that this failure refl…
Dual-Pathway Circuits of Object Hallucination in Vision-Language Models
Jiaxin Liu, Ding Zhong, Yue Wang +6
Vision-language models (VLMs) have demonstrated remarkable capabilities in bridging visual perception and natural language understanding, enabling a wide range of multimodal reason…
A Tale of Two Variances: When Single-Seed Benchmarks Fail in Bayesian Deep Learning
Qishi Zhan, Minxuan Hu, Liang He +2
In limited-data settings, a single endpoint mean of an evaluation metric such as the Continuous Ranked Probability Score (CRPS) is itself a random variable, yet it is routinely rep…
Unstable Rankings in Bayesian Deep Learning Evaluation
Qishi Zhan, Minxuan Hu, Guansu Wang +2
Standard evaluations of Bayesian deep learning methods assume that metric estimates are reliable, but we show this assumption fails under data scarcity. Method rankings are not onl…
Assessment of Generative Named Entity Recognition in the Era of Large Language Models
Qi Zhan, Yile Wang, Hui Huang
Named entity recognition (NER) is evolving from a sequence labeling task into a generative paradigm with the rise of large language models (LLMs). We conduct a systematic evaluatio…
How Order-Sensitive Are LLMs? OrderProbe for Deterministic Structural Reconstruction
Zhaolu Kang, Yingjie He, Kehan Jiang +18
Large language models (LLMs) excel at semantic understanding, yet their ability to reconstruct internal structure from scrambled inputs remains underexplored. Sentence-level restor…