INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

tics

-0.71

loads

-0.70

 Fathers

-0.69

earth

-0.67

baugh

-0.67

ebook

-0.66

 Clouds

-0.63

 acad

-0.63

dos

-0.63

 Scholars

-0.62

POSITIVE LOGITS

outine

0.82

uzzle

0.72

 corridor

0.71

incinn

0.70

emic

0.68

vernment

0.68

redd

0.65

vg

0.65

UST

0.64

ļéĨĴ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.