google/gemma-3-12b-it Sandwich Benchmark Breakdown

Photo-by-photo

Model Results

Breaking down how close the model answered each question, compared to humans.

Photo 01Bacon Lettuce Tomato

google/gemma-3-12b-it

100.0% yes0.0% no

Gap3.7%

Model readLeans yes

Photo 02Dodge Van

google/gemma-3-12b-it

0.0% yes100.0% no

Gap7.0%

Model readLeans no

Photo 03Sub Sandwich

google/gemma-3-12b-it

100.0% yes0.0% no

Gap5.5%

Model readLeans yes

Photo 04Sandwich Costume

google/gemma-3-12b-it

0.0% yes100.0% no

Gap40.9%

Model readLeans no

Photo 05Grilled Cheese

google/gemma-3-12b-it

100.0% yes0.0% no

Gap4.4%

Model readLeans yes

Photo 06Grilled Cheese Pineapple

google/gemma-3-12b-it

100.0% yes0.0% no

Gap8.3%

Model readLeans yes

Photo 07Kitten in Bread

google/gemma-3-12b-it

0.0% yes100.0% no

Gap54.2%

Model readLeans no

Photo 08Hamburger

google/gemma-3-12b-it

99.0% yes1.0% no

Gap26.0%

Model readLeans yes

Photo 09Hashbrown Sandwich

google/gemma-3-12b-it

100.0% yes0.0% no

Gap40.6%

Model readLeans yes

Photo 10Hot Dog

google/gemma-3-12b-it

0.0% yes100.0% no

Gap39.8%

Model readLeans no

Photo 11Pickle Sandwich

google/gemma-3-12b-it

85.0% yes15.0% no

Gap19.4%

Model readLeans yes

Photo 12Avocado Tea

google/gemma-3-12b-it

100.0% yes0.0% no

Gap7.2%

Model readLeans yes

Photo 13Panini

google/gemma-3-12b-it

100.0% yes0.0% no

Gap7.6%

Model readLeans yes

Photo 14Cookie PB

google/gemma-3-12b-it

39.0% yes61.0% no

Gap12.5%

Model readLeans no

Photo 15Chicken Wrap

google/gemma-3-12b-it

51.0% yes49.0% no

Gap28.4%

Model readLeans yes

Photo 16Waffle Ice Cream

google/gemma-3-12b-it

18.0% yes82.0% no

Gap48.3%

Model readLeans no

Photo 17Sloppy Joe

google/gemma-3-12b-it

100.0% yes0.0% no

Gap20.6%

Model readLeans yes

Photo 18Cigarette Sandwich

google/gemma-3-12b-it

0.0% yes100.0% no

Gap29.8%

Model readLeans no

Photo 19KFC Double Down

google/gemma-3-12b-it

100.0% yes0.0% no

Gap44.3%

Model readLeans yes

Photo 20Bagel PB&J

google/gemma-3-12b-it

100.0% yes0.0% no

Gap53.4%

Model readLeans yes

Photo	Vote SplitHuman response	Gap	Read
Photo 01Bacon Lettuce Tomato	google/gemma-3-12b-it 100.0% yes0.0% no	3.7%absolute gap	Leans yesPeople mostly said yes
Photo 02Dodge Van	google/gemma-3-12b-it 0.0% yes100.0% no	7.0%absolute gap	Leans noPeople mostly said no
Photo 03Sub Sandwich	google/gemma-3-12b-it 100.0% yes0.0% no	5.5%absolute gap	Leans yesPeople mostly said yes
Photo 04Sandwich Costume	google/gemma-3-12b-it 0.0% yes100.0% no	40.9%absolute gap	Leans noHuman knife-edge
Photo 05Grilled Cheese	google/gemma-3-12b-it 100.0% yes0.0% no	4.4%absolute gap	Leans yesPeople mostly said yes
Photo 06Grilled Cheese Pineapple	google/gemma-3-12b-it 100.0% yes0.0% no	8.3%absolute gap	Leans yesPeople mostly said yes
Photo 07Kitten in Bread	google/gemma-3-12b-it 0.0% yes100.0% no	54.2%absolute gap	Leans noHuman knife-edge
Photo 08Hamburger	google/gemma-3-12b-it 99.0% yes1.0% no	26.0%absolute gap	Leans yesSplit concept
Photo 09Hashbrown Sandwich	google/gemma-3-12b-it 100.0% yes0.0% no	40.6%absolute gap	Leans yesHuman knife-edge
Photo 10Hot Dog	google/gemma-3-12b-it 0.0% yes100.0% no	39.8%absolute gap	Leans noSplit concept
Photo 11Pickle Sandwich	google/gemma-3-12b-it 85.0% yes15.0% no	19.4%absolute gap	Leans yesSplit concept
Photo 12Avocado Tea	google/gemma-3-12b-it 100.0% yes0.0% no	7.2%absolute gap	Leans yesPeople mostly said yes
Photo 13Panini	google/gemma-3-12b-it 100.0% yes0.0% no	7.6%absolute gap	Leans yesPeople mostly said yes
Photo 14Cookie PB	google/gemma-3-12b-it 39.0% yes61.0% no	12.5%absolute gap	Leans noHuman knife-edge
Photo 15Chicken Wrap	google/gemma-3-12b-it 51.0% yes49.0% no	28.4%absolute gap	Leans yesSplit concept
Photo 16Waffle Ice Cream	google/gemma-3-12b-it 18.0% yes82.0% no	48.3%absolute gap	Leans noSplit concept
Photo 17Sloppy Joe	google/gemma-3-12b-it 100.0% yes0.0% no	20.6%absolute gap	Leans yesSplit concept
Photo 18Cigarette Sandwich	google/gemma-3-12b-it 0.0% yes100.0% no	29.8%absolute gap	Leans noSplit concept
Photo 19KFC Double Down	google/gemma-3-12b-it 100.0% yes0.0% no	44.3%absolute gap	Leans yesHuman knife-edge
Photo 20Bagel PB&J	google/gemma-3-12b-it 100.0% yes0.0% no	53.4%absolute gap	Leans yesHuman knife-edge