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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 gobl

-0.73

oun

-0.71

romeda

-0.70

 destro

-0.68

 existed

-0.66

mort

-0.64

 Checking

-0.62

walk

-0.61

 parted

-0.61

warts

-0.61

POSITIVE LOGITS

 Ging

0.73

0.70

)+

0.64

 Shap

0.63

XP

0.62

âĢ¢âĢ¢

0.62

 pads

0.61

)",

0.60

 Cancer

0.60

 Cheong

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.