-
UC San Diego @icebergnlp SVCL@UCSD
- San Diego, CA
- https://kishorepv.github.io
Stars
We tested Jev, Laya, and Claude to see how they stacked up when it came to configuring a system.
Delving deep into Engagement Prediction of Short Videos (ECCV 2024)
Scored tooling katas, habit rules, and pre-baked warm-start sessions for LLM coding agents
Pocket Flow: Codebase to Tutorial
Skills for Real Engineers. Straight from my .agents directory.
Official repository of the paper "MuQ: Self-Supervised Music Representation Learning with Mel Residual Vector Quantization".
Open-Source Information Retrieval Courses @ TU Wien
An extensive and commented list of resources on Late-Interaction Multivector Retrieval.
A playable terminal FPS whose simulation and framebuffer are calculated in SQLite.
PyTorch-Lightning Library for Neural News Recommendation
A low code Machine Learning personalized ranking service for articles, listings, search results, recommendations that boosts user engagement. A friendly Learn-to-Rank engine
Example to help understand how the BM25 term based ranking model works in search applications
Deprecated historical repo. Superlinked now develops SIE, a self-hosted inference engine for embeddings, reranking, OCR, extraction, and document processing.
MTEB: State-of-the-art evaluation of embeddings across languages and modalities
build your own vector database -- the littlest hnsw
All parts of Claude Code's system prompt, builtin tool descriptions, subagent prompts (Plan/Explore/Task), and utility prompts. Updated for each Claude Code version.
A Toolbox for MultiModal Recommendation. Integrating 10+ Models...
[WWW 2025] The dataset of the paper "A Large-scale Dataset with Behavior, Attributes, and Content of Mobile Short-video Platform"
Tokenization and analysis pipeline for full-text search
Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
An agentic skills framework & software development methodology that works.
The hyperteacher, autoteaching in a metaharness.
Claude Code learns from your corrections: self-correcting memory that compounds over 50+ sessions. Context engineering, parallel worktrees, agent teams, and 17 battle-tested skills.