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

 Cron

-0.72

egal

-0.68

 summ

-0.67

enberg

-0.65

 regime

-0.61

sche

-0.60

ignment

-0.60

uting

-0.59

 conc

-0.58

jection

-0.58

POSITIVE LOGITS

 Roses

0.76

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.76

ICO

0.72

 guiActiveUnfocused

0.72

ãĥīãĥ©

0.71

eatures

0.71

razil

0.70

éļ

0.70

ãĥ©ãĥ³

0.70

ãĤ¤ãĥĪ

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.