INDEX

Explanations

references to quantities or totals followed by numerical values

oai_token-act-pair · gpt-3.5-turbo

phrases that quantify or specify numerical totals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tera

-0.70

pring

-0.63

nesses

-0.63

 Presence

-0.62

ness

-0.56

pod

-0.56

era

-0.55

 condem

-0.55

ity

-0.54

 Males

-0.53

POSITIVE LOGITS

 eight

0.99

six

0.94

 seven

0.93

 thirteen

0.93

 nine

0.91

 THREE

0.89

 fourteen

0.88

 nineteen

0.86

 four

0.86

 eighteen

0.85

Activations Density 0.046%

references to quantities or totals followed by numerical values

phrases that quantify or specify numerical totals

No Comments

No Known Activations