INDEX

Explanations

phrases or sentences indicating coercion, influence, or control over others

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate coercion or manipulation

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aml

-0.69

las

-0.62

lished

-0.62

Dub

-0.61

asus

-0.59

 Garage

-0.57

 Cannes

-0.57

framework

-0.57

©¶æ¥µ

-0.55

ĳå£«

-0.55

POSITIVE LOGITS

 themselves

1.59

 selves

1.13

selves

1.08

 individually

0.93

 their

0.82

 THEIR

0.81

 careers

0.80

 behave

0.74

 mouths

0.73

 collectively

0.70

Activations Density 0.773%

phrases or sentences indicating coercion, influence, or control over others

phrases that indicate coercion or manipulation

No Comments

No Known Activations