INDEX

Explanations

ASCII characters commonly used in Eastern European languages

oai_token-act-pair · gpt-3.5-turbo

the question mark symbol

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 handshake

-0.77

Downloadha

-0.76

sterdam

-0.75

 combination

-0.74

 shape

-0.73

chnology

-0.73

ovie

-0.72

 rotating

-0.71

 revolving

-0.71

 deaf

-0.70

POSITIVE LOGITS

¿

1.39

į

1.16

Ð¾Ð

1.13

ĭ

1.12

Ðµ

1.11

ł

1.07

»

1.06

ÑĢ

1.05

¼

1.05

Ĭ

1.05

Activations Density 0.004%

ASCII characters commonly used in Eastern European languages

the question mark symbol

No Comments

No Known Activations