INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

entious

-0.82

reading

-0.79

OND

-0.69

atur

-0.69

pb

-0.68

idon

-0.67

PB

-0.67

ammy

-0.66

MIC

-0.65

FUL

-0.65

POSITIVE LOGITS

theless

0.95

 approaches

0.73

igham

0.73

 delic

0.66

 behaviors

0.64

 combatants

0.64

 decorations

0.64

 behavi

0.63

 possibilities

0.63

 agre

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.