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

Å

-0.82

Phi

-0.78

plate

-0.76

 Mens

-0.72

otyp

-0.71

 COUR

-0.70

Times

-0.66

otypes

-0.65

Ily

-0.64

owicz

-0.60

POSITIVE LOGITS

 actionGroup

0.79

 dissolve

0.63

Sport

0.63

 favour

0.62

ahoo

0.62

vez

0.62

 favor

0.61

 silenced

0.61

mask

0.60

 shout

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.