INDEX

Explanations

expressions related to public opinion or reactions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating public sentiment or collective behaviors related to people

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

 srfAttach

-0.97

Accessory

-0.80

Sales

-0.73

hent

-0.70

è£ħ

-0.69

inventoryQuantity

-0.69

TX

-0.67

Unix

-0.67

enegger

-0.66

idy

-0.65

POSITIVE LOGITS

 alike

0.74

 hating

0.74

 confuse

0.71

 starving

0.70

 whom

0.68

 addicted

0.68

 misunderstand

0.66

 themselves

0.66

 smarter

0.66

igans

0.65

Activations Density 0.359%

expressions related to public opinion or reactions

phrases indicating public sentiment or collective behaviors related to people

No Comments

No Known Activations