INDEX

Explanations

Spa-related terms

oai_token-act-pair · gpt-3.5-turbo

references to spas and related establishments

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Carbuncle

-0.83

vernment

-0.76

ĪĴ

-0.72

soType

-0.71

 Bulldogs

-0.70

carry

-0.70

 eleph

-0.70

ç¥ŀ

-0.69

orate

-0.68

 headlined

-0.68

POSITIVE LOGITS

iken

0.83

Hut

0.81

isen

0.79

Tus

0.78

aji

0.76

 Tanz

0.73

yrim

0.72

uten

0.71

inen

0.70

Kok

0.68

Activations Density 0.031%

Spa-related terms

references to spas and related establishments

No Comments

No Known Activations