INDEX

Explanations

instances where individuals or entities are defending or protecting themselves

oai_token-act-pair · gpt-3.5-turbo

phrases that involve defending or separating from others

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

heny

-0.68

 Fulton

-0.66

 Syndicate

-0.65

 Rail

-0.65

Mil

-0.65

 Onion

-0.64

 Lyons

-0.64

ammy

-0.63

 Federation

-0.63

 Derby

-0.62

POSITIVE LOGITS

selves

1.22

 tremend

0.85

 selves

0.80

 underwater

0.79

 withd

0.77

 creatively

0.77

 spontaneously

0.76

 submar

0.73

pecially

0.72

 exting

0.72

Activations Density 0.039%

instances where individuals or entities are defending or protecting themselves

phrases that involve defending or separating from others

No Comments

No Known Activations