INDEX

Explanations

phrases related to perceptions of intelligence or competence in relation to social and political issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

SharedCtor

-0.64

 Numerade

-0.56

AndEndTag

-0.55

ثيق

-0.54

 TestBed

-0.52

DUT

-0.49

bors

-0.49

 SHOWING

-0.48

 recognising

-0.48

 <<<<<<<<<<<<<<

-0.47

POSITIVE LOGITS

 utafitiHapana

0.66

 somehow

0.61

 صوتيه

0.59

 magically

0.56

stateProvider

0.55

 ویکی‌پدیا

0.49

 Universitaria

0.49

 tropicales

0.49

raszamy

0.49

 economica

0.48

Activations Density 0.180%

phrases related to perceptions of intelligence or competence in relation to social and political issues

No Comments

No Known Activations