INDEX

Explanations

references to award-winning achievements

oai_token-act-pair · gpt-3.5-turbo

phrases related to award-winning accolades and achievements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

agara

-0.75

matic

-0.72

matter

-0.69

abus

-0.69

onomic

-0.68

ocene

-0.68

omach

-0.67

alle

-0.66

olen

-0.65

umo

-0.64

POSITIVE LOGITS

 actress

0.81

 Pulitzer

0.81

 novelist

0.81

 author

0.78

 bestselling

0.77

AFTA

0.74

seller

0.72

lisher

0.71

 heroine

0.69

 journalist

0.68

Activations Density 0.085%

references to award-winning achievements

phrases related to award-winning accolades and achievements

No Comments

No Known Activations