INDEX

Explanations

proper nouns related to a specific person or location, potentially related to news and politics

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ston" as part of names or terms

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

glers

-0.75

ä»£

-0.70

giving

-0.70

 mileage

-0.69

Pwr

-0.67

TIT

-0.67

 STATS

-0.66

lete

-0.66

Magikarp

-0.65

metic

-0.64

POSITIVE LOGITS

ston

0.95

 Churchill

0.93

ews

0.89

enberg

0.80

ian

0.79

igan

0.78

igans

0.78

ishing

0.77

ium

0.77

ese

0.76

Activations Density 0.017%

proper nouns related to a specific person or location, potentially related to news and politics

occurrences of the word "ston" as part of names or terms

No Comments

No Known Activations