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

©¶æ

-0.82

uminati

-0.80

 VMware

-0.78

 Leilan

-0.75

Jes

-0.74

 Worse

-0.73

 Zoro

-0.69

TeX

-0.69

reon

-0.68

POSITIVE LOGITS

achusetts

0.86

onduct

0.78

glomer

0.77

UCK

0.75

aughter

0.73

igration

0.69

ucky

0.68

ablished

0.68

icity

0.68

igr

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.