INDEX

Explanations

instances where people are taking action collectively or individually

oai_token-act-pair · gpt-3.5-turbo

phrases indicating action or participation by people in events or activities

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

 Advertisement

-0.63

 Increases

-0.60

externalActionCode

-0.60

 Seller

-0.54

 Sprite

-0.52

berus

-0.49

 Allows

-0.48

stery

-0.48

 Lavrov

-0.48

 Drawing

-0.48

POSITIVE LOGITS

onite

0.75

 themselves

0.67

 theirs

0.65

 whom

0.62

 converge

0.60

 interchangeable

0.57

 respectively

0.57

 alike

0.56

ebted

0.55

 collectively

0.55

Activations Density 0.992%

instances where people are taking action collectively or individually

phrases indicating action or participation by people in events or activities

No Comments

No Known Activations