INDEX

Explanations

locations or venues, particularly related to sports events like "Old Trafford" and "Anfield"

oai_token-act-pair · gpt-3.5-turbo

references to specific football stadiums

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Medicaid

-0.77

berman

-0.77

 Labrador

-0.75

 Revised

-0.68

eric

-0.67

ARB

-0.64

audi

-0.64

Operation

-0.63

Nek

-0.62

 SNAP

-0.62

POSITIVE LOGITS

 Trafford

0.96

 Stadium

0.87

minster

0.86

 Sporting

0.84

iership

0.83

bledon

0.82

 Wembley

0.80

 stadiums

0.79

 Dock

0.79

 stadium

0.78

Activations Density 0.019%

locations or venues, particularly related to sports events like "Old Trafford" and "Anfield"

references to specific football stadiums

No Comments

No Known Activations