INDEX

Explanations

references to relationships and familial connections

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

foreign technical terms or descriptions

np_acts-logits-general · gemini-2.5-flash-lite

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

featureID

-0.50

RectangleBorder

-0.44

 BoxDecoration

-0.43

StructEnd

-0.43

ScopeManager

-0.42

 للاسماء

-0.41

LEncoder

-0.40

nologue

-0.39

addCriterion

-0.38

 تضيفلها

-0.38

POSITIVE LOGITS

TestingModule

0.51

 picioare

0.49

AutoScaleMode

0.48

 unconsciously

0.48

 dreapta

0.46

 involuntarily

0.44

 NgModule

0.42

 unwittingly

0.42

marino

0.41

出版年

0.41

Activations Density 0.127%

references to relationships and familial connections

foreign technical terms or descriptions

No Comments

No Known Activations