INDEX

Explanations

units of measurement related to area

oai_token-act-pair · gpt-3.5-turbo

measurements of area

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Kern

-0.77

 bearer

-0.76

 Morg

-0.67

 Pioneer

-0.66

 Easter

-0.66

 guiActiveUnfocused

-0.66

roy

-0.65

 Johann

-0.64

 Kavanaugh

-0.64

 Fargo

-0.64

POSITIVE LOGITS

Sov

0.95

ļ

0.95

inct

0.93

ittal

0.92

tain

0.87

iencies

0.87

urst

0.86

izoph

0.85

ighter

0.85

reen

0.85

Activations Density 0.041%

units of measurement related to area

measurements of area

No Comments

No Known Activations