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

undai

-0.86

hook

-0.80

duty

-0.79

tes

-0.77

©¶æ¥µ

-0.77

sonian

-0.76

script

-0.76

atory

-0.74

egal

-0.72

ignt

-0.72

POSITIVE LOGITS

 Worst

0.72

 Dying

0.71

SPONSORED

0.68

PDATE

0.65

 Passage

0.65

 Patron

0.63

 Brav

0.62

CIS

0.62

 Conditions

0.60

IRE

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.