INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

AddHtmlAttribute

-0.71

 חיצוניים

-0.66

zzleHttp

-0.65

 snippetHide

-0.62

ReusableCell

-0.60

 HttpNotFound

-0.59

PerformLayout

-0.59

 InputDecoration

-0.57

 Chwiliwch

-0.57

 BorderSide

-0.56

POSITIVE LOGITS

 Pristup

0.38

 apropiado

0.38

…

0.37

 adecuado

0.37

 buenas

0.35

 regionales

0.35

 ezen

0.34

turnstile

0.34

 negotiators

0.34

 adecuados

0.34

Activations Density 0.000%

No Known Activations

This feature has no known activations.