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

Hug

-0.94

Pyr

-0.81

 Solitaire

-0.73

 Pick

-0.73

 Pound

-0.73

Pon

-0.68

 Marshal

-0.68

 Buddy

-0.67

 Crown

-0.67

 Letter

-0.66

POSITIVE LOGITS

ancouver

0.89

undo

0.85

luent

0.84

intent

0.83

stanbul

0.83

imates

0.81

icago

0.79

ngth

0.77

ashington

0.76

lda

0.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.