INDEX

Explanations

references to the concept of "whole" or "entire" entities

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

whole foods and tissues

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

featureID

-0.61

Aps

-0.52

sedi

-0.50

bsen

-0.48

 bridesmaid

-0.43

 slipper

-0.43

reys

-0.43

TTS

-0.42

matology

-0.42

 Minerva

-0.42

POSITIVE LOGITS

Whole

1.46

 Whole

1.35

whole

1.25

 WHOLE

1.16

 whole

1.13

 ganze

0.68

 whol

0.65

 entier

0.60

 inteiro

0.60

Hole

0.57

Activations Density 0.011%

references to the concept of "whole" or "entire" entities

whole foods and tissues

No Comments

No Known Activations