INDEX

Explanations

Did-questions, or questions beginning with the word "Did"

oai_token-act-pair · gpt-3.5-turbo

questions beginning with "Did."

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

 strengths

-0.67

endiary

-0.60

 supported

-0.59

 accommod

-0.58

 nearer

-0.57

 tied

-0.56

Appearance

-0.55

 shortest

-0.54

isSpecialOrderable

-0.53

 accounted

-0.53

POSITIVE LOGITS

1.05

actic

1.02

you

1.00

 anyone

0.94

 anybody

0.92

ier

0.84

nt

0.82

 somebody

0.79

YOU

0.75

 someone

0.75

Activations Density 0.051%

Did-questions, or questions beginning with the word "Did"

questions beginning with "Did."

No Comments

No Known Activations