INDEX

Explanations

locations and centers of significance or concentration

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

central places and hubs

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

 préf

-0.40

 version

-0.35

odw

-0.33

gra

-0.33

 modified

-0.33

-0.32

 versión

-0.32

 Alemania

-0.32

Computed

-0.32

 sentence

-0.31

POSITIVE LOGITS

featureID

0.79

 EconPapers

0.78

 &___

0.76

Portail

0.75

 centres

0.72

 hubs

0.72

IntoConstraints

0.71

parsedMessage

0.71

ніципалі

0.70

 سكانية

0.70

Activations Density 0.014%

locations and centers of significance or concentration

central places and hubs

No Comments

No Known Activations