INDEX

Explanations

phrases related to strong disapproval or criticism

oai_token-act-pair · gpt-3.5-turbo

instances of condemnation related to various actions or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

OTE

-0.73

IER

-0.69

soDeliveryDate

-0.66

aldo

-0.66

timer

-0.65

 Surviv

-0.64

icle

-0.64

llular

-0.63

enture

-0.63

emis

-0.63

POSITIVE LOGITS

 abuses

0.99

 atrocities

0.95

 racism

0.93

 vandalism

0.91

 bigotry

0.88

 hypocrisy

0.88

 abuse

0.88

enance

0.86

 violence

0.84

 actions

0.84

Activations Density 0.083%

phrases related to strong disapproval or criticism

instances of condemnation related to various actions or events

No Comments

No Known Activations