Unverified50% confidenceFactExact time
MaRN is an open-source library built on PyTorch, with the core idea of training neural networks through low-dimensional parameter mapping.
1
Sources
50%
Confidence
Long-term
Relevance
10/9/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedNeuralPGN is a local-first, open-source chess opening trainer that uses the FSRS spaced repetition algorithm and is built on the Tauri framework69% similarUnverifiedMaRN constrains or reparameterizes network parameter updates into a lower-dimensional subspace, thereby changing the optimization dynamics and resource consumption patterns of training.68% similarUnverified基于梯度的可解释性工具在PyTorch等训练框架中更早成熟，而在纯推理引擎（如llama.cpp）中实现难度更高67% similarUnverifiedFor deep reinforcement learning scenarios, RL3 reconstructs the environment on the server side using the PyBullet physics engine and provides the PPO algorithm for training.64% similarUnverifiedThe pretraining and fine-tuning paradigm originated from NLP and computer vision, with its core being large-scale data pretraining to learn general low-level features, and downstream tasks adapted using a small amount of labeled data.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/991074API
curl https://kongchang.com/api/v1/knowledge/claims/991074MCP
get_claim(id=991074)