INDEX

Explanations

references to physical properties like density

oai_token-act-pair · gpt-3.5-turbo

mentions of "density"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

opal

-0.82

uberty

-0.80

unes

-0.80

akia

-0.77

porary

-0.75

soDeliveryDate

-0.73

 WATCHED

-0.73

bara

-0.72

oÄŁ

-0.70

Ago

-0.69

POSITIVE LOGITS

 density

1.04

 ratio

0.88

density

0.84

 ratios

0.83

 cooker

0.80

 dens

0.75

 eater

0.74

 utilization

0.74

 gradient

0.73

 clust

0.73

Activations Density 0.009%

references to physical properties like density

mentions of "density"

No Comments

No Known Activations