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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Obj

-0.73

ipop

-0.63

syn

-0.63

Â¨

-0.61

 Blossom

-0.60

 pron

-0.60

 Firm

-0.59

 Slowly

-0.59

rim

-0.58

isha

-0.58

POSITIVE LOGITS

isSpecialOrderable

0.73

 McGregor

0.73

BILITIES

0.69

ãĤ¦ãĤ¹

0.68

atever

0.67

theless

0.66

anchester

0.63

Lua

0.63

externalActionCode

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.