INDEX

Explanations

mentions of or related to public transport users

oai_token-act-pair · gpt-3.5-turbo

references to transit riders and their experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

aples

-0.73

argon

-0.67

orial

-0.66

ilts

-0.65

mberg

-0.65

 Crimea

-0.65

ritz

-0.64

 Bhar

-0.64

gur

-0.63

 Indonesian

-0.61

POSITIVE LOGITS

hip

1.31

 riders

0.99

 rider

0.88

bike

0.88

enty

0.86

pool

0.82

accompan

0.77

comed

0.76

mount

0.75

chairs

0.74

Activations Density 0.015%

mentions of or related to public transport users

references to transit riders and their experiences

No Comments

No Known Activations