INDEX

Explanations

words starting with "un" that depict a negative connotation

oai_token-act-pair · gpt-3.5-turbo

prefixes that convey negation or absence

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

 enhancement

-0.71

 Mechdragon

-0.66

 splits

-0.66

Duo

-0.65

 Madden

-0.64

 Maxwell

-0.64

 edge

-0.64

 periphery

-0.64

 appellant

-0.63

 portions

-0.63

POSITIVE LOGITS

fortunately

1.55

usual

1.50

necessary

1.44

acceptable

1.42

foreseen

1.41

fair

1.41

pleasant

1.40

expected

1.38

mask

1.38

sett

1.38

Activations Density 0.021%

words starting with "un" that depict a negative connotation

prefixes that convey negation or absence

No Comments

No Known Activations