INDEX

Explanations

proper nouns related to political figures, universities, and events

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and locations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

>>>>>>>>

-0.76

 Portug

-0.71

 Chains

-0.70

 Debor

-0.67

 NEWS

-0.62

 Episcopal

-0.62

 apocalypse

-0.61

afety

-0.61

hr

-0.59

 Claus

-0.59

POSITIVE LOGITS

amel

0.91

alos

0.90

doms

0.86

jet

0.86

insky

0.85

aji

0.84

endor

0.83

arios

0.83

nit

0.81

deen

0.80

Activations Density 0.023%

proper nouns related to political figures, universities, and events

proper nouns, particularly names and locations

No Comments

No Known Activations