4 4 13

Josef Chen PRO

josefchen

https://josef.cn

josefchen

AI & ML interests

Food

Recent Activity

authored a paper 1 day ago

When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models

liked a Space 1 day ago

josefchen/orchestration-is-allocation

new activity 1 day ago

josefchen/co-failure-67-models:Add task category to metadata

View all activity

Organizations

Papers 7

spaces 2

Combining LLMs Rarely Beats the Single Best Model

🎲

beta=P(all wrong): the co-failure ceiling on LLM ensembles

The Physical AI Inference Gap in Batch-1 LLM Decode

🪜

Interactive companion to the batch-1 LLM decode paper

models 5

Josef Chen PRO

AI & ML interests

Recent Activity

Organizations

Papers 7

spaces 2

Combining LLMs Rarely Beats the Single Best Model

The Physical AI Inference Gap in Batch-1 LLM Decode

models 5

josefchen/smolvla-libero-obj-t5-lora

josefchen/smolvla-libero-obj-t1-lora

josefchen/smolvla-libero-obj-t8-lora

josefchen/smolvla-libero-obj-t2-lora

josefchen/smolvla-libero-obj-t6-lora

datasets 1

josefchen/co-failure-67-models

Josef Chen PRO

AI & ML interests

Recent Activity

Organizations

Papers 7

spaces 2 Sort: Recently updated

Combining LLMs Rarely Beats the Single Best Model

The Physical AI Inference Gap in Batch-1 LLM Decode

models 5 Sort: Recently updated

datasets 1

spaces 2

models 5