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

agen

-0.80

 Speedway

-0.67

ruciating

-0.67

Ga

-0.66

 grid

-0.65

 Frie

-0.64

 Gazette

-0.64

rongh

-0.63

ãĥ¯ãĥ³

-0.63

ako

-0.63

POSITIVE LOGITS

 concess

0.71

 sleeper

0.66

 Communism

0.64

 defe

0.63

::::::::

0.59

 deserving

0.59

Recent

0.59

ternity

0.59

Tip

0.59

 induction

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.