INDEX

Explanations

personal pronouns and self-references

oai_token-act-pair · gpt-3.5-turbo

instances of the pronoun "I" and expressions of personal reflection or experience

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

¿½

-0.63

 Plaint

-0.63

ļéĨĴ

-0.63

%%%%

-0.57

ä¹ĭ

-0.56

Appearances

-0.55

Weak

-0.54

rules

-0.54

 harms

-0.54

Children

-0.53

POSITIVE LOGITS

'm

1.36

 figured

1.23

've

1.09

 decided

1.09

 stumbled

1.08

 realised

1.01

 thought

1.00

am

1.00

 realized

0.97

 guess

0.96

Activations Density 0.155%

personal pronouns and self-references

instances of the pronoun "I" and expressions of personal reflection or experience

No Comments

No Known Activations