INDEX

Explanations

terms related to military or authority figures

oai_token-act-pair · gpt-3.5-turbo

terms related to military or organizational command structures

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

 advoc

-0.76

shine

-0.66

Woman

-0.65

 Trout

-0.64

employment

-0.64

Labor

-0.64

Gamer

-0.62

Bloomberg

-0.61

Enough

-0.60

Wan

-0.59

POSITIVE LOGITS

eering

1.37

eers

1.18

eer

1.09

ments

1.00

nance

0.93

ant

0.89

ment

0.88

antine

0.87

quartered

0.86

aining

0.85

Activations Density 0.036%

terms related to military or authority figures

terms related to military or organizational command structures

No Comments

No Known Activations