INDEX

Explanations

phrases related to negative attributes or criticism

oai_token-act-pair · gpt-3.5-turbo

phrases that express strong negative sentiments or criticisms

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

ESE

-0.87

cade

-0.83

uese

-0.82

ynthesis

-0.81

aea

-0.76

maxwell

-0.75

imum

-0.75

 Details

-0.73

assadors

-0.73

iets

-0.72

POSITIVE LOGITS

 irresponsible

1.43

 discriminatory

1.41

 immoral

1.39

 abusive

1.38

 counterproductive

1.37

 wasteful

1.37

 unethical

1.36

 hypocritical

1.35

 disrespectful

1.35

 unconstitutional

1.31

Activations Density 0.141%

phrases related to negative attributes or criticism

phrases that express strong negative sentiments or criticisms

No Comments

No Known Activations