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

-|

-0.72

<?

-0.68

agate

-0.66

endon

-0.64

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

-0.62

rett

-0.62

*/

-0.60

::::::::

-0.60

Copy

-0.60

eh

-0.59

POSITIVE LOGITS

wives

0.70

regon

0.67

aukee

0.66

 stabilization

0.66

VIDIA

0.65

 Flint

0.64

agonists

0.61

lication

0.60

itage

0.60

Wil

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.