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

 behavi

-0.84

soDeliveryDate

-0.84

PDATE

-0.76

uliffe

-0.76

izoph

-0.75

henko

-0.75

ymm

-0.74

odynam

-0.73

 eleph

-0.72

nyder

-0.71

POSITIVE LOGITS

:]

0.75

 Contra

0.69

 coli

0.67

 presses

0.67

iHUD

0.65

den

0.64

pak

0.64

 Libre

0.64

 Barbarian

0.63

row

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.