INDEX

Explanations

names, surnames, and related words that potentially refer to people

oai_token-act-pair · gpt-3.5-turbo

the name "Mil" and its variations in different contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

*/(

-0.66

ĪĴ

-0.65

 reconciliation

-0.63

 skelet

-0.62

yright

-0.62

|--

-0.60

 Genocide

-0.59

 concess

-0.59

bye

-0.59

 Archdemon

-0.58

POSITIVE LOGITS

veland

0.70

outh

0.68

 Gaal

0.66

enta

0.65

andre

0.64

utsche

0.64

irts

0.64

west

0.63

otte

0.63

igham

0.63

Activations Density 0.191%

names, surnames, and related words that potentially refer to people

the name "Mil" and its variations in different contexts

No Comments

No Known Activations