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

 kasarigan

-1.06

ConstraintMaker

-1.05

IntoConstraints

-0.94

__(/*!

-0.87

 nahilalakip

-0.86



-0.84

 betweenstory

-0.82

<pad>

-0.81

<unused8>

-0.80

[@BOS@]

-0.79

POSITIVE LOGITS

↵↵

0.53

CrossRef

0.51

<eos>

0.45

0.39

↵

0.38

0.35

0.33

 Even

0.31

].

0.30

Activations Density 0.000%

No Known Activations

This feature has no known activations.