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

romeda

-0.74

ģ«

-0.70

 habitable

-0.70

ected

-0.70

 inheritance

-0.67

ĪĴ

-0.67

stellar

-0.67

ior

-0.63

erva

-0.62

 tear

-0.62

POSITIVE LOGITS

IJ

0.95

sic

0.75

 Jagu

0.75

see

0.74

 Lauder

0.72

again

0.71

tti

0.71

Que

0.70

according

0.69

among

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.