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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 aloud

-0.71

 baptized

-0.70

 Chal

-0.70

 reluct

-0.66

 Rahman

-0.64

 chorus

-0.63

 ladder

-0.62

 railing

-0.62

 stages

-0.61

 Parliament

-0.60

POSITIVE LOGITS

roma

0.79

vae

0.74

ulf

0.72

wana

0.72

ahu

0.72

origin

0.70

inous

0.68

hyd

0.67

itle

0.67

Pack

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.