INDEX

Explanations

proper nouns and titles

oai_token-act-pair · gpt-3.5-turbo

mentions of key political figures or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

cellence

-0.59

 Errors

-0.54

à¨

-0.53

Joined

-0.53

iuses

-0.52

="#

-0.51

found

-0.50

 defective

-0.50

 manic

-0.50

ãĢĲ

-0.50

POSITIVE LOGITS

 will

1.08

 likely

1.03

TBA

1.00

 preferably

0.99

 would

0.97

 feasibility

0.96

 ideally

0.94

TBD

0.94

likely

0.93

will

0.90

Activations Density 1.209%

proper nouns and titles

mentions of key political figures or events

No Comments

No Known Activations