INDEX

Explanations

references to justifications or reasons, typically framed as "grounds."

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

phrases related to legal or justifiable reasons

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Sins

-0.71

 Lomb

-0.70

apt

-0.69

 paycheck

-0.66

 centralized

-0.65

thal

-0.65

oser

-0.63

 Abyssal

-0.63

omez

-0.62

 Twins

-0.61

POSITIVE LOGITS

 grounds

1.22

keepers

1.10

keeper

1.03

ground

0.88

keeping

0.88

grounds

0.87

GROUND

0.81

ãĤ³

0.80

 Grounds

0.79

æĦ

0.74

Activations Density 0.015%

references to justifications or reasons, typically framed as "grounds."

phrases related to legal or justifiable reasons

No Comments

No Known Activations