INDEX

Explanations

phrases related to decision-making and approval

oai_token-act-pair · gpt-3.5-turbo

expressions of potential outcomes or alternatives

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

)</

-0.73

estern

-0.62

but

-0.57

heric

-0.56

?",

-0.55

*)

-0.55

ursday

-0.52

erent

-0.51

asus

-0.51

leted

-0.51

POSITIVE LOGITS

yet

1.13

yet

0.97

nor

0.86

 anymore

0.84

 anyway

0.79

 either

0.77

 unless

0.77

 anytime

0.75

 whatsoever

0.72

Activations Density 0.599%

phrases related to decision-making and approval

expressions of potential outcomes or alternatives

No Comments

No Known Activations