๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Zhuofeng Li
@zhuofengli96475
RSI. SR @GoogleDeepMind / PhD Student @TAMU. Former @StanfordAILab | TIGER-Lab @UWaterloo. Lead OpenRSI-Index, AgentFlow, Verl-Tool, OpenResearcher, DCI.
๊ฐ€์ž… January 2024
648 ํŒ”๋กœ์ž‰ ์ค‘    679 ํŒฌ
๐Ÿš€ DR-DCI improves raw DCI from 62.9% โ†’ 71.2% on BrowseComp-Plus while also reducing ๐Ÿ’ฐ cost and โฑ๏ธ runtime. ๐Ÿ” Retrieval-for-discovery + ๐Ÿ› ๏ธ DCI-for-investigation is a practical decomposition for scaling DCI agents to large-scale corpora (10M). Congrats to the authors! ๐Ÿ‘
๋” ๋ณด๊ธฐ
๐Ÿš€ Introducing DR-DCI: Scaling Direct Corpus Interaction via Dynamic Workspace Expansion Search agents face a basic tradeoff: Retrieval scales to massive corpora, but often exposes evidence only as ranked snippets or bounded document views. Direct Corpus Interaction (DCI) gives agents precise terminal-style control with tools like rg, grep, find, and read, but raw full-corpus search becomes slow and unstable as the corpus grows. DR-DCI bridges the two. The agent dynamically calls pull(query, topK) to expand a local workspace, then performs DCI inside that bounded workspace for evidence search, comparison, and verification. Retriever = corpus-level discovery DCI = workspace-level investigation Result: on BrowseComp-Plus, DR-DCI improves Raw DCI from 62.9% to 71.2%, while reducing wall time and estimated cost. With workspace-preserving context reset, it reaches 73.25%. HF paper: Code: More details below ๐Ÿ‘‡
๋” ๋ณด๊ธฐ