INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 accord

-0.67

WTO

-0.63

Amb

-0.62

 appar

-0.62

 Creed

-0.61

 RCMP

-0.61

 scramble

-0.61

 theorem

-0.61

 reckoning

-0.60

 Mahar

-0.60

POSITIVE LOGITS

avorite

0.91

assin

0.80

nesday

0.79

eki

0.79

okia

0.76

ucket

0.76

imore

0.75

obi

0.74

arnaev

0.74

aco

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.