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

ipal

-0.81

misc

-0.77

NetMessage

-0.71

rum

-0.70

yth

-0.70

 clashed

-0.66

vre

-0.65

icon

-0.64

 decom

-0.63

 misc

-0.63

POSITIVE LOGITS

 mileage

0.75

 Countdown

0.69

 encouragement

0.67

 Stop

0.66

 chapters

0.65

keye

0.64

eers

0.63

 Haku

0.61

 Step

0.61

 Answer

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.