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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atown

-0.92

culated

-0.80

culation

-0.78

oming

-0.72

veland

-0.72

resp

-0.71

igree

-0.71

ENCY

-0.70

reditary

-0.70

culus

-0.67

POSITIVE LOGITS

nas

0.73

etta

0.71

letters

0.66

fold

0.66

 Nikola

0.63

na

0.63

âĻ¥

0.63

letter

0.62

plugins

0.60

 overlook

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.