INDEX

Explanations

comparative phrases indicating one thing being significantly greater than another

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that indicate something is larger or longer

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

istries

-0.71

Winged

-0.69

efe

-0.65

resa

-0.65

Constructed

-0.63

ached

-0.63

umbn

-0.63

eni

-0.63

ãĥ´

-0.62

acho

-0.62

POSITIVE LOGITS

 usual

1.40

 anticipated

1.18

 expected

1.13

 ours

0.97

 advertised

0.91

 previously

0.89

 others

0.89

usual

0.89

 anything

0.88

 typical

0.88

Activations Density 0.080%

comparative phrases indicating one thing being significantly greater than another

comparative phrases that indicate something is larger or longer

No Comments

No Known Activations