INDEX

Explanations

phrases where something is described as the best or hardest part

oai_token-act-pair · gpt-3.5-turbo

statements about significant or notable aspects of experiences or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Klux

-0.80

¬¼

-0.73

 millenn

-0.72

incinn

-0.71

apons

-0.71

ãĤī

-0.71

 confir

-0.71

 practition

-0.70

ħĭ

-0.70

elsius

-0.68

POSITIVE LOGITS

ioned

0.83

 Myster

0.80

ials

0.75

 Payton

0.73

icularly

0.71

ially

0.67

icular

0.67

uncture

0.67

 Darrell

0.67

icle

0.66

Activations Density 0.027%

phrases where something is described as the best or hardest part

statements about significant or notable aspects of experiences or events

No Comments

No Known Activations