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

olate

-0.80

querade

-0.79

cible

-0.79

attr

-0.76

ifestyle

-0.72

olated

-0.70

atures

-0.68

umn

-0.65

 insepar

-0.65

selves

-0.65

POSITIVE LOGITS

Reloaded

0.77

Rod

0.72

wig

0.71

 Krish

0.70

 clerics

0.70

acters

0.69

ĳå£«

0.67

ãĥ¯

0.67

bard

0.64

ãĢĮ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.