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

unt

-0.66

 scrut

-0.62

UI

-0.61

 quotas

-0.61

OU

-0.59

å¯

-0.59

Dialog

-0.59

_>

-0.58

VR

-0.58

::::::::

-0.58

POSITIVE LOGITS

ĸļ

0.87

ronic

0.76

©¶æ

0.72

collar

0.71

 Flavoring

0.71

aine

0.67

mble

0.66

antha

0.65

raviolet

0.64

emade

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.