Cross-Pollination Retrieval: Finding Transferable Ideas Across Fields in Scientific Literature
We study cross-pollination retrieval, the problem of identifying transferable ideas across research fields, and introduce CPR-Dataset, a large-scale dataset constructed by mining cross-topic citations.
ProofPaper ↗
Key points
- Many scientific advances come from cross-pollination, carrying an idea from one field over to a problem in another.
- Yet retrieving papers across fields remains underexplored, with existing benchmarks confined to a narrow range of fields.
- Its evaluation split CPR-eval has 1,300 queries in 26 fields, and its training split CPR-train has 10,573 pairs.
- The gain carries over to domains held out from training and to analogy and reasoning-intensive retrieval.
Sources (1)
- [1]Cross-Pollination Retrieval: Finding Transferable Ideas Across Fields in Scientific LiteraturearXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 19, 02:29 AM
We study cross-pollination retrieval, the problem of identifying transferable ideas across research fields, and introduce CPR-Dataset, a large-scale dataset constructed by mining cross-topic citations.
Many scientific advances come from cross-pollination, carrying an idea from one field over to a problem in another.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 19, 2026TorchNeuron: running HF models on AWS Trainium is easier than ever
- Oct 18, 2026Comfy-Org/Fibo-2.0
- Oct 18, 2026Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark: DeepSeek-V4.1-Flash-1017 EXL3 2.9bpw on 2x NVIDIA DGX Spark (GB10) with TensorFold: 2.1-2.3x the vLLM kit's decode speed.
- Oct 17, 2026AdamPM: Memory-Efficient LLM Optimization via Moment Pooling
- Oct 16, 2026Hugging Face ❤️ TorchTPU: TPUs, Natively
- Oct 16, 2026[Research] 2,788 tokens/s on a single RTX 5090 with a diffusion LLM — Introducing ORL (fully open source)