INDEX

Explanations

strong positive descriptions of experiences or objects

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

positive sentiment and appealing qualities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tatuagens

-0.40

astanza

-0.39

ساز

-0.37

Jody

-0.37

 UNHCR

-0.37

 obrazu

-0.37

 keju

-0.36

futbol

-0.36

inergic

-0.36

webElementXpaths

-0.36

POSITIVE LOGITS

 gorgeous

0.47

Gorgeous

0.44

 beautiful

0.44

 irresistible

0.44

lovely

0.42

éez

0.42

DockStyle

0.42

 ſte

0.42

 Gorgeous

0.42

simply

0.41

Activations Density 0.030%

strong positive descriptions of experiences or objects

positive sentiment and appealing qualities

No Comments

No Known Activations