INDEX

Explanations

mentions of social issues and policies involving people in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to individuals or groups within social contexts

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

 Lank

-0.73

 srfAttach

-0.73

inventoryQuantity

-0.70

Sil

-0.67

Dad

-0.65

Mas

-0.64

NES

-0.64

BALL

-0.64

Appearance

-0.64

heny

-0.62

POSITIVE LOGITS

 smugglers

1.23

 traffickers

0.96

who

0.92

folk

0.86

hood

0.81

 smuggling

0.79

 donating

0.76

 affected

0.71

 born

0.71

 vaccinated

0.70

Activations Density 0.124%

mentions of social issues and policies involving people in various contexts

references to individuals or groups within social contexts

No Comments

No Known Activations