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

 Elect

-0.65

 resid

-0.62

 Martian

-0.59

oscopic

-0.58

 compositions

-0.58

 Covenant

-0.58

 Georgian

-0.57

icent

-0.57

 Moon

-0.57

 electron

-0.56

POSITIVE LOGITS

ington

0.94

 news

0.85

atche

0.80

ewitness

0.76

isine

0.74

eka

0.70

iku

0.68

veland

0.68

********************************

0.67

âĺħâĺħ

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.