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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aler

-0.75

vy

-0.74

uce

-0.74

ering

-0.73

utic

-0.72

omsky

-0.70

é¾įå

-0.69

annels

-0.69

ansk

-0.69

utics

-0.69

POSITIVE LOGITS

interstitial

0.86

etheless

0.84

ously

0.75

Magikarp

0.67

 newsp

0.67

HIP

0.65

 induct

0.64

 simul

0.64

ially

0.63

 seizure

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.