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

displayText

-0.82

 Tanz

-0.79

 Sark

-0.71

 Bosh

-0.68

Mub

-0.67

Ö¼

-0.67

 Nare

-0.66

 dishes

-0.66

 Bake

-0.66

 Garn

-0.65

POSITIVE LOGITS

docs

0.84

resp

0.69

RAW

0.69

mys

0.68

ammed

0.66

OO

0.64

wik

0.64

WC

0.63

Ec

0.63

 Editors

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.