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

lyn

-0.77

 Sell

-0.75

isson

-0.73

omed

-0.68

 Seller

-0.67

zan

-0.66

ĸļ

-0.65

oked

-0.63

ulf

-0.63

inventory

-0.62

POSITIVE LOGITS

gib

0.73

ESA

0.66

_-

0.65

VICE

0.63

OPLE

0.62

phrine

0.60

atism

0.59

 blindness

0.58

image

0.57

 livest

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.