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

évaluateur

-0.64

ftagPool

-0.62

extAlignment

-0.52

Referanser

-0.51

bindingNavigator

-0.51

balleur

-0.51

 PLWABN

-0.49

 interviewers

-0.48

 interviewer

-0.48

insics

-0.48

POSITIVE LOGITS

 privés

3.06

 privé

1.58

 privées

1.55

 privados

1.46

 privée

1.38

 privadas

1.34

 private

1.33

 privado

1.33

 privada

1.29

 Private

1.19

Activations Density 0.000%

No Known Activations

This feature has no known activations.