INDEX

Explanations

No Explanations Found

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

 polymorph

-0.79

ĪĴ

-0.75

 helic

-0.75

 sqor

-0.71

 lamb

-0.67

 reciproc

-0.67

 microsc

-0.67

 somehow

-0.66

 branching

-0.65

vow

-0.65

POSITIVE LOGITS

olitan

0.89

rique

0.80

eenth

0.77

suit

0.73

uum

0.72

cled

0.71

rived

0.71

--------------------------------------------------------

0.71

isodes

0.70

sie

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.