INDEX

Explanations

words related to the color white followed by a number, which may indicate a level of emphasis or intensity

oai_token-act-pair · gpt-3.5-turbo

references to the concept of whiteness and its related social implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

REM

-0.88

yrinth

-0.84

cffffcc

-0.83

ally

-0.82

tis

-0.80

olls

-0.75

vati

-0.72

ategory

-0.72

nel

-0.72

interstitial

-0.72

POSITIVE LOGITS

 supremacist

1.49

 supremacists

1.40

 supremacy

1.25

 suprem

1.10

 nationalist

1.06

 nationalists

0.98

bread

0.97

 phosphorus

0.97

 collar

0.95

caps

0.94

Activations Density 0.044%

words related to the color white followed by a number, which may indicate a level of emphasis or intensity

references to the concept of whiteness and its related social implications

No Comments

No Known Activations