INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

velt

-0.69

alore

-0.68

htaking

-0.67

nce

-0.65

mite

-0.64

reme

-0.62

wolves

-0.61

ignty

-0.61

thing

-0.61

omorph

-0.60

POSITIVE LOGITS

IFT

0.82

ARC

0.70

Management

0.68

 Walton

0.66

rift

0.66

NAS

0.65

IELD

0.65

othal

0.64

OUT

0.64

oyd

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.