circuits-llm

Index

AdamWStateCircuit.LLM.Training
adamwStepCircuit.LLM.Training
addGptGradsCircuit.LLM.Backprop
Aff 
1 (Type/Class)Circuit.LLM.SSM
2 (Data Constructor)Circuit.LLM.SSM
affACircuit.LLM.SSM
affAVecCircuit.LLM.SSM
affBCircuit.LLM.SSM
affBVecCircuit.LLM.SSM
affCompCircuit.LLM.SSM
affCompVecCircuit.LLM.SSM
AffVec 
1 (Type/Class)Circuit.LLM.SSM
2 (Data Constructor)Circuit.LLM.SSM
applyMaskCircuit.LLM.Training
argmaxCircuit.LLM.Inference
assocScanCircuit.LLM.SSM
assocScanVecCircuit.LLM.SSM
assocSSMCircuit.LLM.SSM
assocSSMVecCircuit.LLM.SSM
bertBackwardCircuit.LLM.Backprop
bgAttnLnBetaCircuit.LLM.Backprop
bgAttnLnGammaCircuit.LLM.Backprop
bgAttnWkCircuit.LLM.Backprop
bgAttnWoCircuit.LLM.Backprop
bgAttnWqCircuit.LLM.Backprop
bgAttnWvCircuit.LLM.Backprop
bgFfnB1Circuit.LLM.Backprop
bgFfnB2Circuit.LLM.Backprop
bgFfnLnBetaCircuit.LLM.Backprop
bgFfnLnGammaCircuit.LLM.Backprop
bgFfnW1Circuit.LLM.Backprop
bgFfnW2Circuit.LLM.Backprop
BlockGrads 
1 (Type/Class)Circuit.LLM.Backprop
2 (Data Constructor)Circuit.LLM.Backprop
BPEEncoding 
1 (Type/Class)Circuit.LLM.BPE
2 (Data Constructor)Circuit.LLM.BPE
BPEErrorCircuit.LLM.BPE
bpeMaxTokenIdCircuit.LLM.BPE
bpeMergeRulesCircuit.LLM.BPE
BPEModel 
1 (Type/Class)Circuit.LLM.BPE
2 (Data Constructor)Circuit.LLM.BPE
bpeRegexCircuit.LLM.BPE
bpeReverseSpecialCircuit.LLM.BPE
bpeSpecialTokensCircuit.LLM.BPE
bpeVersionCircuit.LLM.BPE
bpeVocabCircuit.LLM.BPE
causalMaskCircuit.LLM.Attention
chunkedLinearAttnCircuit.LLM.Mixer
chunkedScanCircuit.LLM.SSM
coupledMultiHeadSSMSystemCircuit.LLM.SSM
crossEntropyBwdCircuit.LLM.Backprop
crossEntropyLossCircuit.LLM.Training
decodeBPECircuit.LLM.BPE
decodeBPEWithPerfCircuit.LLM.BPE
deltaAttnPrefillCircuit.LLM.Mixer
deltaAttnStepCircuit.LLM.Mixer
eluPlus1Circuit.LLM.Mixer
emptyKvCircuit.LLM.Mixer
emptyLinearCircuit.LLM.Mixer
encodeBPECircuit.LLM.BPE
encodeBPEWithPerfCircuit.LLM.BPE
encodedTokensCircuit.LLM.BPE
FeedForward 
1 (Type/Class)Circuit.LLM.GPT
2 (Data Constructor)Circuit.LLM.GPT
ffB1Circuit.LLM.GPT
ffB2Circuit.LLM.GPT
ffW1Circuit.LLM.GPT
ffW2Circuit.LLM.GPT
forwardCircuit.LLM.GPT
gatedDeltaPrefillCircuit.LLM.Mixer
gatedDeltaStepCircuit.LLM.Mixer
geluCircuit.LLM.GPT
geluBwdCircuit.LLM.Backprop
generateCircuit.LLM.Inference
ggBlocksCircuit.LLM.Backprop
ggHeadCircuit.LLM.Backprop
ggHeadBCircuit.LLM.Backprop
ggLnBetaCircuit.LLM.Backprop
ggLnGammaCircuit.LLM.Backprop
ggWpeCircuit.LLM.Backprop
ggWteCircuit.LLM.Backprop
Gpt 
1 (Type/Class)Circuit.LLM.GPT
2 (Data Constructor)Circuit.LLM.GPT
Gpt2LargeCircuit.LLM.Weights
Gpt2MediumCircuit.LLM.Weights
Gpt2SizeCircuit.LLM.Weights
Gpt2SmallCircuit.LLM.Weights
Gpt2XlCircuit.LLM.Weights
gptBackwardCircuit.LLM.Backprop
gptBlocksCircuit.LLM.GPT
GptConfig 
1 (Type/Class)Circuit.LLM.GPT
2 (Data Constructor)Circuit.LLM.GPT
GptGrads 
1 (Type/Class)Circuit.LLM.Backprop
2 (Data Constructor)Circuit.LLM.Backprop
gptHeadCircuit.LLM.GPT
gptHeadBCircuit.LLM.GPT
gptLnBetaCircuit.LLM.GPT
gptLnGammaCircuit.LLM.GPT
gptNEmbdCircuit.LLM.GPT
gptNHeadCircuit.LLM.GPT
gptNLayerCircuit.LLM.GPT
gptVocabSizeCircuit.LLM.GPT
gptWpeCircuit.LLM.GPT
gptWteCircuit.LLM.GPT
greedySampleCircuit.LLM.Inference
initAdamWCircuit.LLM.Training
InvalidTokenIdCircuit.LLM.BPE
kdaLiteStepCircuit.LLM.Mixer
KvCache 
1 (Type/Class)Circuit.LLM.Mixer
2 (Data Constructor)Circuit.LLM.Mixer
kvKCircuit.LLM.Mixer
kvVCircuit.LLM.Mixer
lastRowCircuit.LLM.Inference
layerNormCircuit.LLM.GPT
layerNormBwdCircuit.LLM.Backprop
linearAttnPrefillCircuit.LLM.Mixer
linearAttnStepCircuit.LLM.Mixer
linearBwdCircuit.LLM.Backprop
LinearState 
1 (Type/Class)Circuit.LLM.Mixer
2 (Data Constructor)Circuit.LLM.Mixer
linSCircuit.LLM.Mixer
linZCircuit.LLM.Mixer
loadBPEModelCircuit.LLM.BPE
loadBPEModelWithPerfCircuit.LLM.BPE
loadGpt2Circuit.LLM.Weights
loadGpt2WithCircuit.LLM.Weights
loadMatrixCircuit.LLM.Weights
loadVectorCircuit.LLM.Weights
maskedCrossEntropyBwdCircuit.LLM.Backprop
maskPositionsCircuit.LLM.Training
mergeHeadsCircuit.LLM.Attention
ModelParseErrorCircuit.LLM.BPE
multiHeadAttentionCircuit.LLM.Attention
multiHeadSSMSystemCircuit.LLM.SSM
numChunksCircuit.LLM.BPE
originalTextCircuit.LLM.BPE
prettifyBPEModelCircuit.LLM.BPE
prettifyEncodingCircuit.LLM.BPE
RegexCompileErrorCircuit.LLM.BPE
runMultiHeadSSMSystemCircuit.LLM.SSM
runSharedInputMultiHeadSSMSystemCircuit.LLM.SSM
runSystemCircuit.LLM.SSM
scaledDotProductAttentionCircuit.LLM.Attention
scaleGptGradsCircuit.LLM.Backprop
scanMixerCircuit.LLM.Mixer
seqSSMCircuit.LLM.SSM
seqSSMVecCircuit.LLM.SSM
sizeConfigCircuit.LLM.Weights
softmaxCircuit.LLM.Attention
softmaxAttnPrefillCircuit.LLM.Mixer
softmaxAttnStepCircuit.LLM.Mixer
softmaxBwdCircuit.LLM.Backprop
softmaxVCircuit.LLM.Inference
splitHeadsCircuit.LLM.Attention
ssmProcessCircuit.LLM.SSM
ssmSystemCircuit.LLM.SSM
ssmSystemVecCircuit.LLM.SSM
tbAttnLnBetaCircuit.LLM.GPT
tbAttnLnGammaCircuit.LLM.GPT
tbAttnWkCircuit.LLM.GPT
tbAttnWoCircuit.LLM.GPT
tbAttnWqCircuit.LLM.GPT
tbAttnWvCircuit.LLM.GPT
tbFfnCircuit.LLM.GPT
tbFfnLnBetaCircuit.LLM.GPT
tbFfnLnGammaCircuit.LLM.GPT
temperatureSampleCircuit.LLM.Inference
topKSampleCircuit.LLM.Inference
trainLoopCircuit.LLM.Training
trainLoopMaskedCircuit.LLM.Training
trainStepCircuit.LLM.Training
trainStepMaskedCircuit.LLM.Training
TransformerBlock 
1 (Type/Class)Circuit.LLM.GPT
2 (Data Constructor)Circuit.LLM.GPT
transformerBlockCircuit.LLM.GPT
updateMatrixCircuit.LLM.Training
updateVectorCircuit.LLM.Training
zeroGptGradsCircuit.LLM.Backprop