INDEX

Explanations

references to prominent historical figures and their relationships

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

historical figures and names

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

 EconPapers

-0.70

InitVars

-0.63

Tembelea

-0.60

OGND

-0.59

wieś

-0.59

forChild

-0.56

hasMoreElements

-0.55

IUrlHelper

-0.55

Erreferentziak

-0.52

devamını

-0.52

POSITIVE LOGITS

AppMethodBeat

0.40

 internetowa

0.32

 meli

0.31

createTextNode

0.29

колко

0.28

úrate

0.28

 patata

0.28

 Malin

0.28

IfExists

0.28

 للمعارف

0.28

Activations Density 0.048%

references to prominent historical figures and their relationships

historical figures and names

No Comments

No Known Activations