INDEX

Explanations

words related to hiding or concealing something

oai_token-act-pair · gpt-3.5-turbo

references to hiding or concealing aspects of identity, truth, or personal issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 roundup

-0.91

rano

-0.67

icion

-0.66

Reviewed

-0.65

unia

-0.64

umar

-0.64

usa

-0.64

sta

-0.63

izo

-0.62

tackle

-0.62

POSITIVE LOGITS

 wrongdoing

1.07

 secrets

1.06

 truths

0.95

 identity

0.94

 identities

0.94

 whereabouts

0.89

 traces

0.84

 vulnerabilities

0.83

 existence

0.82

 differences

0.79

Activations Density 0.265%

words related to hiding or concealing something

references to hiding or concealing aspects of identity, truth, or personal issues

No Comments

No Known Activations