INDEX

Explanations

references to directories and file structures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

+#+

-0.47

Personendaten

-0.45

findpost

-0.44

 InputDecoration

-0.42

 llorar

-0.41

?>>

-0.38

исленность

-0.38

TemporalType

-0.38

ActionPerformed

-0.38

VPB

-0.37

POSITIVE LOGITS

ViewFeatures

0.60

 itſelf

0.53

 purpoſe

0.44

 myſelf

0.43

 incendie

0.43

AutoScaleMode

0.42

 intStringLen

0.42

 queſta

0.42

 يتيمه

0.41

lectricité

0.40

Activations Density 0.018%

references to directories and file structures

No Comments

No Known Activations