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

Zup

-0.79

jee

-0.72

 Cuomo

-0.71

ZIP

-0.69

angelo

-0.68

 Albany

-0.65

 Superman

-0.65

 Saras

-0.65

EMOTE

-0.65

 Sidd

-0.64

POSITIVE LOGITS

ensation

0.79

ivation

0.76

ounds

0.76

uay

0.75

orescence

0.74

ahon

0.74

ounding

0.72

areth

0.70

arij

0.70

ahah

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.