INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 neighb

-0.72

bowl

-0.69

thur

-0.69

estones

-0.67

 confir

-0.65

 flask

-0.65

 feather

-0.65

iah

-0.64

 helicop

-0.64

 skelet

-0.63

POSITIVE LOGITS

ļéĨĴ

0.86

osterone

0.85

oute

0.84

ption

0.78

Constructed

0.77

othal

0.67

ptive

0.67

MAD

0.66

 Madness

0.66

ulator

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.