INDEX

Explanations

terms and phrases related to research, analysis, and formal procedures

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

Tembelea

-0.59

TagMode

-0.45

 resol

-0.44

jsxFileName

-0.43

pis

-0.41

 Identity

-0.41

cá

-0.40

__':

-0.40

界的

-0.39

 awareness

-0.39

POSITIVE LOGITS

 Normdatei

0.52

 <>",

0.48

 demografica

0.45

asteroide

0.43

twimg

0.43

ulemon

0.43

parsedMessage

0.41

 iNdEx

0.41

 policiales

0.40

 Exacts

0.39

Activations Density 0.082%

terms and phrases related to research, analysis, and formal procedures

No Comments

No Known Activations