INDEX

Explanations

exclamatory statements or emotional expressions

oai_token-act-pair · gpt-3.5-turbo

repeated use of a specific character or symbol in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

 cloak

-0.69

wom

-0.69

 sled

-0.69

 shroud

-0.68

 coats

-0.65

 protective

-0.65

 Shroud

-0.64

 pace

-0.63

 sidx

-0.63

SAR

-0.63

POSITIVE LOGITS

ľ

1.58

¬

1.48

¼

1.27

½

1.26

ł

1.22

º

1.22

Ń

1.15

Ł

1.14

¿

1.12

«

1.11

Activations Density 0.240%

exclamatory statements or emotional expressions

repeated use of a specific character or symbol in the text

No Comments

No Known Activations