INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

taboola

-0.73

 bots

-0.72

Ana

-0.67

staff

-0.65

Marie

-0.65

 ACTIONS

-0.65

Posts

-0.64

translation

-0.64

Admin

-0.64

 Psycho

-0.63

POSITIVE LOGITS

thur

0.76

terday

0.69

fired

0.68

brush

0.66

cil

0.66

 perman

0.66

 Antiqu

0.64

 shortcut

0.63

 retention

0.62

sworth

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.