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

stock

-0.73

ebted

-0.72

ForgeModLoader

-0.70

stocks

-0.70

 "$:/

-0.69

stan

-0.67

maxwell

-0.67

wart

-0.65

ethy

-0.62

tch

-0.62

POSITIVE LOGITS

liest

0.78

cule

0.67

rocal

0.67

bidden

0.66

KNOWN

0.63

icles

0.63

oret

0.63

tical

0.63

 Odin

0.62

LESS

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.