INDEX

Explanations

texts discussing potential or possibilities

oai_token-act-pair · gpt-3.5-turbo

references to "potential" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ogie

-0.80

ĪĴ

-0.80

bows

-0.79

cipline

-0.72

ograph

-0.72

cloth

-0.70

llan

-0.69

chet

-0.68

oak

-0.68

OTO

-0.68

POSITIVE LOGITS

izons

0.89

ibilities

0.84

 usefulness

0.84

 implications

0.83

ities

0.80

 future

0.79

 adversaries

0.78

 payoff

0.76

 unintended

0.76

atility

0.74

Activations Density 0.021%

texts discussing potential or possibilities

references to "potential" in various contexts

No Comments

No Known Activations