INDEX

Explanations

locations or streets in different cities

oai_token-act-pair · gpt-3.5-turbo

specific locations or addresses in a context of events or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 conferences

-0.74

 countries

-0.74

 treaties

-0.72

 nations

-0.72

 domestically

-0.70

 physiology

-0.69

 portfolios

-0.68

 prem

-0.67

 democracies

-0.66

 abroad

-0.65

POSITIVE LOGITS

Police

1.00

 sidx

0.96

 PHOTO

0.93

police

0.87

Neigh

0.85

hello

0.84

andals

0.84

OSED

0.83

 photographed

0.83

 Residents

0.83

Activations Density 0.471%

locations or streets in different cities

specific locations or addresses in a context of events or actions

No Comments

No Known Activations