INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 FANTASY

-0.72

missible

-0.69

ertodd

-0.69

Appearances

-0.68

 Investig

-0.68

 Forensic

-0.67

 Aliens

-0.67

ACTION

-0.65

Quality

-0.65

 Paran

-0.65

POSITIVE LOGITS

 wills

0.70

 wishes

0.66

 lane

0.64

 jokes

0.64

 neutrality

0.64

 wrath

0.64

 loans

0.63

 caucuses

0.63

 masc

0.63

kind

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.