INDEX

Explanations

contractions of "are not"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "aren't" in various contexts

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

lag

-0.69

ramid

-0.66

town

-0.65

ocaust

-0.64

cision

-0.64

 residue

-0.63

oter

-0.62

owan

-0.61

cycle

-0.61

growth

-0.60

POSITIVE LOGITS

't

1.05

ajor

0.83

Marginal

0.73

 gonna

0.68

nces

0.66

NOT

0.65

nos

0.65

 wont

0.63

nai

0.61

nell

0.61

Activations Density 0.013%

contractions of "are not"

instances of the word "aren't" in various contexts

No Comments

No Known Activations