INDEX

Explanations

expressions of strong emotions or beliefs

oai_token-act-pair · gpt-3.5-turbo

phrases that express different forms of opinion or perspective

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

bid

-0.80

cas

-0.70

bats

-0.68

head

-0.66

fn

-0.66

idity

-0.65

notations

-0.64

uity

-0.63

heads

-0.63

orgetown

-0.62

POSITIVE LOGITS

 sorts

1.25

 pure

0.82

 theirs

0.79

icial

0.71

 ours

0.70

 genuine

0.70

 necessity

0.67

Lear

0.66

 honour

0.65

aeper

0.65

Activations Density 0.218%

expressions of strong emotions or beliefs

phrases that express different forms of opinion or perspective

No Comments

No Known Activations