INDEX

Explanations

references to the city of Hong Kong

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Hong Kong."

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

nesota

-0.73

 placeholder

-0.69

 inference

-0.68

ally

-0.66

CTR

-0.65

 enhancement

-0.64

 wedge

-0.63

uality

-0.63

igm

-0.62

 Gemini

-0.61

POSITIVE LOGITS

 Kong

1.38

awei

1.13

 Hong

0.92

nam

0.91

Hong

0.89

wei

0.87

jin

0.87

zhou

0.87

moon

0.83

ju

0.82

Activations Density 0.010%

references to the city of Hong Kong

occurrences of the term "Hong Kong."

No Comments

No Known Activations