INDEX

Explanations

mentions of drinking water

oai_token-act-pair · gpt-3.5-turbo

mentions of drinking water or alcoholic beverages

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

eq

-0.77

orp

-0.74

debian

-0.70

elin

-0.69

roe

-0.67

 Kinnikuman

-0.67

moving

-0.66

sure

-0.65

sha

-0.65

Egypt

-0.64

POSITIVE LOGITS

 drinking

1.11

cohol

1.07

 alcohol

1.02

 beverage

1.00

 beverages

0.99

 drinkers

0.98

 drink

0.95

 drinks

0.93

 Drink

0.90

 whiskey

0.89

Activations Density 0.010%

mentions of drinking water

mentions of drinking water or alcoholic beverages

No Comments

No Known Activations