INDEX

Explanations

the letter 'u' followed by a number indicating intensity

oai_token-act-pair · gpt-3.5-turbo

instances of the letter 'u' in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Noir

-0.76

DonaldTrump

-0.72

 Cove

-0.70

CPC

-0.70

 Preferred

-0.69

Missing

-0.69

 Citation

-0.68

enburg

-0.66

 Paddock

-0.66

 Dickinson

-0.64

POSITIVE LOGITS

pperc

1.20

mpire

1.07

zbek

0.95

reth

0.92

prising

0.89

seless

0.88

raviolet

0.85

du

0.84

pport

0.82

berman

0.81

Activations Density 0.009%

the letter 'u' followed by a number indicating intensity

instances of the letter 'u' in various contexts

No Comments

No Known Activations