INDEX

Explanations

references to specific locations and geographic places

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kysy

-0.44

rø

-0.36

 tuot

-0.35

 Autorisations

-0.35

OUGHT

-0.34

parametrize

-0.34

Fø

-0.34

sum

-0.34

indd

-0.33

 CHAPITRE

-0.33

POSITIVE LOGITS

 annually

0.52

 festi

0.51

 jährlich

0.50

rawDesc

0.48

RenderAtEndOf

0.47

每年

0.46

 المعيارى

0.46

Celebrate

0.45

 celebrate

0.44

anthene

0.44

Activations Density 0.009%

references to specific locations and geographic places

No Comments

No Known Activations