Unverified50% confidenceFactExact time
The forward computation of selective scan is a recurrence along the sequence dimension: h_t = A_bar_t · h_{t-1} + B_bar_t · x_t, where A_bar and B_bar are determined by Δ.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedNumerical gradient checking approximates gradients using the central difference formula (f(θ+ε) - f(θ-ε)) / (2ε) and compares them with analytical gradients; a relative error below 1e-5 is considered correct, while an error above 1e-3 almost certainly indicates a bug.57% similarUnverifiedAfter unifying the top-k sweep from the author's previous post onto the same 9 questions, the curve flattened between 0.67 and 0.83, with all seven runs falling within the judge noise band, so the sweep could not distinguish among the k values; the earlier conclusion of 'optimal k=8' does not hold.56% similarUnverifiedNumerical gradient checking approximates the derivative via finite differences: for each parameter θ, compute [L(θ+ε) - L(θ-ε)] / (2ε), where ε is a very small number (such as 1e-4).56% similarVerifiedThe formula for a cubic Bézier curve is B(t) = (1-t)³P₀ + 3(1-t)²tP₁ + 3(1-t)t²P₂ + t³P₃, where t∈[0,1] is the parameter and P₀ through P₃ are four control points54% similarUnverifiedMultiplying the value of the least significant bit obtained from x & (-x) by the 32-bit de Bruijn constant 0x077CB531 and then right-shifting by 27 bits yields a 5-bit index that uniquely corresponds to a bit position in a precomputed lookup table.53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1070371API
curl https://kongchang.com/api/v1/knowledge/claims/1070371MCP
get_claim(id=1070371)