INDEX

Explanations

comparisons or contrasts between different subjects or entities

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

comparative phrases indicating superiority or emphasis

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

Constructed

-0.81

alities

-0.72

 largeDownload

-0.71

rain

-0.68

english

-0.66

*/(

-0.66

imens

-0.64

nels

-0.63

orthy

-0.63

ãĤ¨

-0.63

POSITIVE LOGITS

 ever

1.19

 usual

1.14

 anything

1.13

 meets

0.99

 anticipated

0.93

 anyone

0.93

any

0.92

 anybody

0.91

 mere

0.81

we

0.80

Activations Density 0.095%

comparisons or contrasts between different subjects or entities

comparative phrases indicating superiority or emphasis

No Comments

No Known Activations