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

 destro

-0.83

 hurd

-0.67

bud

-0.66

 toile

-0.63

TeX

-0.62

tro

-0.62

 sentimental

-0.61

 tong

-0.60

ptions

-0.60

 passionately

-0.60

POSITIVE LOGITS

ussion

0.71

json

0.69

 Imran

0.68

phabet

0.67

byter

0.65

APTER

0.65

ictionary

0.64

aredevil

0.64

dm

0.62

RAW

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.