INDEX

Explanations

phrases related to a variety or diversity of items or concepts

oai_token-act-pair · gpt-3.5-turbo

references to various categories or types

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

edia

-0.88

aeper

-0.76

instein

-0.71

edi

-0.71

Ö¼

-0.70

ainer

-0.67

opsis

-0.64

 Ronaldo

-0.63

rican

-0.62

gary

-0.62

POSITIVE LOGITS

 imaginable

0.79

itionally

0.70

hell

0.69

cape

0.66

 combinations

0.65

entially

0.65

 converge

0.65

wcsstore

0.65

atisf

0.64

adoes

0.64

Activations Density 0.027%

phrases related to a variety or diversity of items or concepts

references to various categories or types

No Comments

No Known Activations