INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Bullets

-0.70

 rationality

-0.65

 egalitarian

-0.60

 reconcil

-0.60

 life

-0.59

 solved

-0.59

 libertarian

-0.58

 Indonesia

-0.58

 abiding

-0.58

haw

-0.57

POSITIVE LOGITS

escent

0.86

resa

0.83

enza

0.82

incial

0.79

SO

0.75

izon

0.75

ãĤ£

0.75

hs

0.74

endez

0.74

asus

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.