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

querque

-0.72

ascus

-0.69

 conclud

-0.69

é¾įå¥ĳå£«

-0.69

 converge

-0.67

 Temper

-0.66

 Bark

-0.64

 guessed

-0.64

 craz

-0.63

ITNESS

-0.62

POSITIVE LOGITS

',

0.73

hof

0.71

amination

0.68

ima

0.68

 Aviation

0.63

'.

0.62

 Publishers

0.62

 Valiant

0.61

 Plugin

0.61

 Secondary

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.