Spaces:
Sleeping
Sleeping
A newer version of the Gradio SDK is available:
6.5.1
metadata
title: PC-Bench
emoji: 📚
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 6.3.0
app_file: app.py
pinned: false
license: mit
short_description: Paper Discovery Benchmark
tags:
- leaderboard
- research
- multi-agent
- paper-retrieval
PC-Bench: Paper Discovery Benchmark
Leaderboard for evaluating AI agents on academic paper retrieval and analysis.
Benchmarks
| Benchmark | Queries | Description |
|---|---|---|
| SemanticBench | 50 | Template-based semantic queries |
| RAbench | 500 | LLM-perturbed natural queries |
Metrics
- MRR - Mean Reciprocal Rank
- R@K - Recall at K (K=1,5,10,20,50)
- Hit Rate - Successful retrieval percentage
Top Results
| Model | Hit Rate | MRR | Time |
|---|---|---|---|
| Qwen3-Coder-30B | 80% | 0.627 | 22s |
| BM25 Baseline | 78% | 0.541 | - |