INDEX

Explanations

comparisons that express superiority or preference against alternatives

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

ContentAsync

-0.70

 ModelExpression

-0.70

featureID

-0.70

 GenerationType

-0.70

FunctionFlags

-0.69

 lenker

-0.69

 Normdatei

-0.69

bootstrapcdn

-0.69

windowFixed

-0.67

Jeografia

-0.67

POSITIVE LOGITS

 previous

0.37

 existing

0.37

 typical

0.35

 traditional

0.34

Typical

0.33

previous

0.32

 prior

0.32

 futile

0.31

 typique

0.31

 confusion

0.30

Activations Density 0.031%

comparisons that express superiority or preference against alternatives

No Comments

No Known Activations