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

ysc

-0.74

ands

-0.72

omaly

-0.72

osc

-0.69

 clock

-0.65

hem

-0.64

 swat

-0.64

oji

-0.64

progress

-0.63

urb

-0.63

POSITIVE LOGITS

 Leone

0.73

Footnote

0.72

 Slaughter

0.67

mma

0.67

âĶĢâĶĢâĶĢâĶĢ

0.62

vine

0.62

felt

0.61

vir

0.61

 Helpful

0.60

ptin

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.