INDEX

Explanations

phrases starting with "Its" describing various entities or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases or terms that indicate official announcements or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 penalty

-0.63

 Ward

-0.63

 prospect

-0.63

 Prospect

-0.62

 Cook

-0.61

 Price

-0.61

 Davis

-0.60

 Center

-0.60

 Sever

-0.59

 front

-0.58

POSITIVE LOGITS

Its

3.40

Pros

1.40

its

1.23

Their

1.18

RIP

1.13

Firstly

1.12

Attempts

1.11

It

1.10

DragonMagazine

1.07

Such

1.04

Activations Density 0.025%

phrases starting with "Its" describing various entities or concepts

phrases or terms that indicate official announcements or statements

No Comments

No Known Activations