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

 Question

-0.73

 renewal

-0.62

 Deity

-0.61

 Advantage

-0.60

 Initi

-0.59

 recharge

-0.59

 reclaim

-0.59

 link

-0.58

 parity

-0.58

 Resource

-0.57

POSITIVE LOGITS

Downloadha

0.89

olar

0.85

ktop

0.84

adoes

0.84

leep

0.82

wagen

0.80

raints

0.80

ocamp

0.79

accompan

0.78

ometimes

0.78

Activations Density 0.000%

No Known Activations

This feature has no known activations.