INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Sands

-0.69

 taxp

-0.69

 tuberculosis

-0.65

 Harding

-0.62

 Norris

-0.61

 ours

-0.61

 Needs

-0.61

 Hurricanes

-0.60

ounty

-0.59

ieves

-0.59

POSITIVE LOGITS

demon

0.85

chrom

0.81

genre

0.79

archive

0.77

channel

0.75

ART

0.74

books

0.74

chem

0.71

videos

0.69

touch

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.