INDEX

Explanations

person names or professions with variations in spelling

oai_token-act-pair · gpt-3.5-turbo

descriptors of characters in narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

lihood

-0.75

 Recommend

-0.70

Recommend

-0.68

 redes

-0.65

 Procedures

-0.64

FTA

-0.63

 Funding

-0.63

 Fiscal

-0.62

HHS

-0.62

 Marketplace

-0.61

POSITIVE LOGITS

ussy

0.87

 blonde

0.77

ampunk

0.77

loving

0.77

airy

0.77

angled

0.76

ppy

0.76

bred

0.75

anky

0.75

 bikini

0.73

Activations Density 0.622%

person names or professions with variations in spelling

descriptors of characters in narratives

No Comments

No Known Activations