INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 uncond

-0.70

Tup

-0.67

om

-0.66

 quietly

-0.63

 precept

-0.62

 silently

-0.62

 Lenin

-0.62

 acutely

-0.61

 territ

-0.61

 mindfulness

-0.60

POSITIVE LOGITS

hire

3.46

ibur

1.10

gery

0.86

BUR

0.84

erie

0.84

haw

0.83

street

0.81

shire

0.81

rity

0.78

orks

0.78

Activations Density 0.000%

No Known Activations

This feature has no known activations.