INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 <>",

-0.44

やっと

-0.40

 efficiency

-0.40

駛

-0.39

 capacity

-0.37

 efficient

-0.37

bi

-0.36

findpost

-0.36

 ConstraintSet

-0.35

Géographie

-0.35

POSITIVE LOGITS

0.56

 fallecido

0.51

0.50

 bénéficiaire

0.50

 condolences

0.49

Obituary

0.48

 fallecimiento

0.48

Portale

0.48

 Condol

0.47

 Obituary

0.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.