INDEX

Explanations

locations specified in a specific format like city and state abbreviations or city and country

oai_token-act-pair · gpt-3.5-turbo

location identifiers, specifically state abbreviations in the United States

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

ufact

-0.84

lihood

-0.82

renheit

-0.75

andowski

-0.71

toggle

-0.71

holding

-0.69

months

-0.68

é¾įå

-0.68

paragraph

-0.67

ror

-0.67

POSITIVE LOGITS

 Dept

0.78

./

0.72

ARA

0.70

FF

0.67

ADA

0.65

 Supervisor

0.64

JR

0.64

FC

0.63

IFF

0.62

BB

0.62

Activations Density 0.037%

locations specified in a specific format like city and state abbreviations or city and country

location identifiers, specifically state abbreviations in the United States

No Comments

No Known Activations