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

Mach

-0.84

bolt

-0.76

Bell

-0.76

vation

-0.74

Mob

-0.72

-0.70

Tube

-0.70

Ru

-0.70

STEM

-0.67

DB

-0.67

POSITIVE LOGITS

0.86

(%

0.76

(<

0.75

(.

0.73

 (âĪĴ

0.72

("

0.70

(=

0.70

(#

0.68

('

0.68

 Hispan

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.