INDEX

Explanations

the word "knew."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "knew," indicating awareness or prior knowledge

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

otion

-0.74

phrine

-0.74

oples

-0.68

 ILCS

-0.68

 rÃ©

-0.68

conservancy

-0.66

olid

-0.66

interstitial

-0.65

ember

-0.65

area

-0.64

POSITIVE LOGITS

 beforehand

0.98

ledged

0.86

lessly

0.78

lege

0.77

ledge

0.77

footed

0.72

saf

0.72

nothing

0.71

cut

0.70

 instinctively

0.70

Activations Density 0.047%

the word "knew."

instances of the word "knew," indicating awareness or prior knowledge

No Comments

No Known Activations