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

scribe

-0.76

soever

-0.74

 molten

-0.69

hig

-0.66

 Swords

-0.66

 fingert

-0.66

arthed

-0.66

uct

-0.65

 livest

-0.64

agra

-0.64

POSITIVE LOGITS

chen

0.66

 courts

0.65

xon

0.65

heny

0.63

 McKenna

0.63

 partners

0.61

onnaissance

0.61

 MacArthur

0.60

SHIP

0.59

 registry

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.