Stories by fzliu

Understanding Warmup-Stable-Decay Learning Rates

Multimodal embeddings for interleaved text and images

What the Hype and Reality of Agents

Multimodal RAG

32K-token context length embedding models

How to win Equalized/Worst positions?

The Modern RAG Stack

OmniParser for Pure Vision Based GUI Agent

Improving Generalization Performance by Switching from Adam to SGD (2017)

Embeddings Drive the Quality of RAG: A Case Study of Chat.LangChain

Rerank-2 and rerank-2-lite: the next generation of Voyage multilingual rerankers

Voyage-3 and voyage-3-lite: small yet mighty general-purpose embedding models

Why opening statistics are hard

Does Refusal Training in LLMs Generalize to the Past Tense?

TxT360

Rerank-2 and rerank-2-lite: the next generation of Voyage multilingual rerankers

32K-token context length embedding models

Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence

How Does a Reranker Work?

Robust (Structured) Text Generation

10
11
12
13
14
15
16
17
18
19