INDEX

Explanations

details related to the city of New Delhi in news articles

oai_token-act-pair · gpt-3.5-turbo

references to specific geographic locations or proper nouns

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

liest

-0.69

Cyp

-0.65

 counties

-0.64

keys

-0.63

 bins

-0.63

 cabinets

-0.61

Oak

-0.60

 Whit

-0.60

 weights

-0.60

 sheets

-0.60

POSITIVE LOGITS

HCR

0.92

ournal

0.91

 Lumpur

0.87

 IMAGES

0.85

ANK

0.81

zsche

0.79

OPE

0.78

AMI

0.77

anmar

0.77

uay

0.76

Activations Density 0.029%

details related to the city of New Delhi in news articles

references to specific geographic locations or proper nouns

No Comments

No Known Activations