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

onom

-0.71

 Videos

-0.70

antle

-0.68

 Personnel

-0.67

rose

-0.65

 Topics

-0.65

Detect

-0.64

ophysical

-0.64

ocracy

-0.63

st

-0.61

POSITIVE LOGITS

 lett

0.68

 rejoice

0.65

 opting

0.63

obyl

0.63

 shelling

0.60

 battle

0.59

issance

0.58

 doorway

0.58

 Piano

0.57

ento

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.