INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

JD

-0.67

 UCHIJ

-0.63

 seeded

-0.62

Quotes

-0.61

 certific

-0.61

 Dayton

-0.58

 subsidies

-0.57

Indiana

-0.56

Plex

-0.56

 certified

-0.55

POSITIVE LOGITS

nir

0.82

 Shang

0.76

icular

0.72

 Lash

0.68

naire

0.68

lier

0.67

 TEXT

0.67

 Trance

0.67

Nib

0.66

odcast

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.