Unverified50% confidenceFactExact time
For a frozen base matrix W0, LoRA represents the learned update as W' = W0 + (alpha/r)BA, where A and B are trainable compact matrices with rank r much smaller than the dimensions of the original matrix.
1
Sources
50%
Confidence
Long-term
Relevance
10/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDuring LoRA training, the pre-trained model weights W₀ are frozen and only the two matrices B and A are trained, with the forward propagation formula H = W₀X + (α/r)BAX.76% similarVerifiedLoRA attaches two low-rank matrices A and B in parallel alongside each target weight matrix, where A is randomly initialized and B is initialized to zero; only these two small matrices are trained, and during inference the product A×B is added back to the original weights.75% similarVerifiedLoRA achieves fine-tuning by inserting low-rank decomposition matrices alongside the model's weight matrices, typically requiring training of only 0.1%–1% of the original parameter count.67% similarUnverifiedLow-rank parameterization decomposes the full covariance matrix into the form Σ = WW^T + D, reducing the parameter count from O(d²) to O(dr).63% similarUnverifiedA confusion matrix is an N×N square matrix where the element in the i-th row and j-th column represents the number of samples whose true class is i but were predicted as j62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/998065API
curl https://kongchang.com/api/v1/knowledge/claims/998065MCP
get_claim(id=998065)