INDEX

Explanations

phrases related to judgments or assessments of situations or conditions

oai_token-act-pair · gpt-3.5-turbo

expressions of perception or opinion

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

arta

-0.85

been

-0.80

tan

-0.75

heed

-0.74

por

-0.72

wall

-0.70

ogie

-0.66

 Yourself

-0.66

veland

-0.66

coin

-0.65

POSITIVE LOGITS

 initially

0.75

 abruptly

0.69

 tremend

0.69

 originally

0.69

nesday

0.68

 seism

0.68

hes

0.65

hers

0.63

 noticeably

0.63

 briefly

0.62

Activations Density 0.428%

phrases related to judgments or assessments of situations or conditions

expressions of perception or opinion

No Comments

No Known Activations