INDEX

Explanations

interrogative sentences that prompt for an explanation or response

oai_token-act-pair · gpt-3.5-turbo

questions or inquiries posed 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

athe

-0.72

 threaded

-0.68

 weld

-0.68

agos

-0.67

 knit

-0.65

 background

-0.64

iannopoulos

-0.63

ema

-0.63

 brim

-0.63

imer

-0.63

POSITIVE LOGITS

 Well

1.52

 Firstly

1.21

 Probably

1.19

Well

1.16

 Quite

1.14

 Plenty

1.11

 Certainly

1.09

 Turns

1.09

 Surely

1.08

 Perhaps

1.08

Activations Density 0.099%

interrogative sentences that prompt for an explanation or response

questions or inquiries posed in the text

No Comments

No Known Activations