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

oola

-0.75

 accuser

-0.69

 anesthesia

-0.69

 linem

-0.68

blast

-0.67

going

-0.65

ittal

-0.63

 sounding

-0.63

 Sorce

-0.61

 billing

-0.61

POSITIVE LOGITS

fect

0.71

ilater

0.68

ris

0.66

ror

0.65

Len

0.64

hap

0.62

iot

0.61

hered

0.60

Hom

0.60

 Hond

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.