INDEX

Explanations

proper nouns and unique terms

oai_token-act-pair · gpt-3.5-turbo

symbols or special characters that represent unique features or themes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

aceous

-0.69

eele

-0.66

 hasht

-0.65

 actor

-0.64

 fodder

-0.63

 Pony

-0.63

oids

-0.62

 stunt

-0.61

 killer

-0.61

 emerging

-0.60

POSITIVE LOGITS

ï¸ı

1.48

ï¸

1.25

ternity

1.13

Â£

1.07

Â¢

1.06

Â¯

1.05

ÃĽ

1.02

Vers

0.94

Cu

0.93

should

0.93

Activations Density 0.020%

proper nouns and unique terms

symbols or special characters that represent unique features or themes

No Comments

No Known Activations