INDEX

Explanations

specific references to locations, especially related to New South Wales (NSW) in Australia

oai_token-act-pair · gpt-3.5-turbo

references to New South Wales (NSW) and associated political matters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãĥł

-0.67

ãĤ¼

-0.64

bows

-0.64

due

-0.64

ror

-0.63

 bearer

-0.61

å£«

-0.61

heimer

-0.61

bow

-0.60

 favoring

-0.59

POSITIVE LOGITS

NSW

1.29

urst

0.87

FW

0.82

ECA

0.75

IRO

0.74

NRL

0.73

ategory

0.73

foundland

0.73

cffff

0.71

 Queensland

0.70

Activations Density 0.009%

specific references to locations, especially related to New South Wales (NSW) in Australia

references to New South Wales (NSW) and associated political matters

No Comments

No Known Activations