INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

kees

-0.94

ette

-0.88

sels

-0.80

istor

-0.75

nor

-0.74

photos

-0.71

oult

-0.71

Unknown

-0.70

shell

-0.70

eta

-0.70

POSITIVE LOGITS

Þ

0.89

 ende

0.77

 carbohyd

0.70

 Goodell

0.67

ß

0.66

nesday

0.66

ð

0.66

beh

0.66

 Licensed

0.65

ACE

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.