INDEX

Explanations

expressions of personal opinions and emotional responses

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

 engraçadas

-0.50

 manguera

-0.46

 CONSTITUTION

-0.41

 BUREAU

-0.41

 Bureau

-0.40

Hentet

-0.40

 valentín

-0.40

 constitutional

-0.40

Pranala

-0.39

 prioridad

-0.39

POSITIVE LOGITS

EndContext

0.59

 myself

0.50

MLLoader

0.46

LLocation

0.43

 pinulongan

0.42

ViewFeatures

0.41

createCanvas

0.40

myself

0.39

styleType

0.39

 gaping

0.39

Activations Density 0.046%

expressions of personal opinions and emotional responses

No Comments

No Known Activations