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

stals

-0.67

âĸ¬

-0.67

sid

-0.65

yx

-0.63

ãĥ¼ãĥĨ

-0.60

 Disp

-0.59

 ferment

-0.59

 Bombs

-0.58

 Democr

-0.58

TX

-0.58

POSITIVE LOGITS

ogn

0.93

ython

0.86

earch

0.83

rupulous

0.76

irus

0.72

ocene

0.70

oneliness

0.70

onductor

0.70

mouth

0.70

odan

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.