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

otaur

-0.75

icter

-0.72

phis

-0.72

agos

-0.67

 Aberdeen

-0.66

utherford

-0.65

odore

-0.65

iets

-0.64

osaurs

-0.64

akin

-0.64

POSITIVE LOGITS

ciplinary

0.83

lip

0.77

actionDate

0.73

Mask

0.68

 marketed

0.68

 polic

0.66

displayText

0.64

pour

0.63

seller

0.62

Agent

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.