Skip to content

Author

Jia-Hao Sun

We have 14 of 60 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Nov 2026

Cortex-FL: Agent-Mediated Collaboration for Federated Learning

Federated learning (FL) enables independent data owners to train models collaboratively without centralizing raw data. However, identifying collaborators, preparing training tasks, and coordinating execution and incentives hinder its practical adoption by organizations without specialized expertise. We present Cortex-F...

Rui Sun, Jin-Feng Peng, Ze-Hua Cheng et al. · 0 citations
Book Open access Nov 2026

In-Context Learning as Approximate Belief Propagation

Few-shot in-context learning (ICL) cannot revise its belief about an early demonstration once a later, contradicting one arrives, so a single misleading example can derail the prediction. Existing remedies are split: retrieval methods choose which examples to include but weight them uniformly; calibration methods adjus...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
Book Open access Nov 2026

Gradient Interaction Graph Optimization

Training large Transformer language models is dominated by the cost of synchronous parameter updates: every block of weights is moved at every step using a gradient computed before any block has moved. The resulting trajectory is wasteful, because per-block updates are evaluated on a frozen snapshot of the rest of the...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
Book Open access Nov 2026

Curvature Graph Mergeability: Predicting Model Compatibility from Per-Layer Hessian Geometry

Curvature Graph Mergeability is introduced, a scoring framework that predicts the interpolation barrier from only cheap, per-layer Hessian curvature features at each endpoint, and proves that the interpolation barrier is upper-bounded by a multilinear function of per-layer curvature mismatch.

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
Book Open access Nov 2026

Cavity-Calibrated Prediction Sets

Cavity-Calibrated Prediction Sets (CCPS) are introduced, which prove distribution-free marginal and joint coverage guarantees and establish a correlation-decay condition under which cavity scores converge to true leave-one-out scores.

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
Book Open access Nov 2026

Cavity-Calibrated Prediction Sets

Conformal prediction equips black-box classifiers with finite-sample, distribution-free coverage guarantees, but its standard application to structured outputs over probabilistic graphical models is hampered by two issues: per-node nonconformity scores derived from approximate beliefs are inflated by self-reinforcing l...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
#small language model Book Open access Nov 2026

Gradient Interaction Graph Optimization

GIGO is introduced, a meta-optimizer that estimates a sparse directed graph of cross-block gradient dependencies online, partitions each step into contiguous sequential stages along this graph, and refreshes downstream gradients using a stale-activation approximation, and is compatible with mixed precision and data par...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
#small language model Book Open access Nov 2026

Bethe Consensus Decoding: Joint Inference on Empirical Pairwise Factor Graphs for Structured LLM Outputs

BCD is introduced, a training-free post-hoc aggregator that reconciles K samples drawn from a frozen LLM by minimising a Bethe free energy on a sparse pairwise factor graph and shows that on tree-structured factor graphs in the population limit, BCD strictly dominates IMV in the induced tree distribution.

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
#graph neural networks Book Open access Nov 2026

Curvature Graph Mergeability: Predicting Model Compatibility from Per-Layer Hessian Geometry

Merging independently trained neural networks --- via weight averaging, model soups, or task-vector arithmetic --- is increasingly central to large-scale model reuse, yet the success of any merge hinges on whether two parameter vectors occupy the same loss basin, a property currently certified only by expensive multi-p...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
Book Open access Sep 2026

Decentralized Learning with Communication-Efficient Learned Gradient Sketches

Decentralized learning eliminates single-point-of-failure risks inherent in parameter-server architectures, but its practical deployment is bottlenecked by the communication cost of exchanging full-precision gradients over peer-to-peer links. Existing compression techniques—random quantization, sparsification, and fixe...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 0 citations
#artificial intelligence Preprint Oct 2026

SoK: Decentralized Agent Economic Infrastructure

This study organizes security and economic requirements into 17 property families over six stages, with receipt soundness and completeness assessed separately, and introduces guarantee closure, a task-relative criterion for determining whether guarantees established at one stage remain available and constrain the later...

Rui Sun, Xi-Han Xiong, Qin Wang et al. · 0 citations

this-that-model-1.0: A typed decision model that decides in 30 ms, for a millionth of a cent

Software delegates more of its branches to models every year: which queue a ticket enters, whether a command is safe to run, whether a claim clears without a person. What the program needs back is not prose. It is one of n declared options and a number it can threshold. Today that costs a round trip to a frontier model...

Ze-Hua Cheng, Wei Dai, Jia-Hao Sun · 10 citations · ⚡3

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.