Scheduled Commit

data/retrieval_battle-f62bcad9-19c4-430b-a7ba-aa2bfeba99dc.jsonl

@@ -79,3 +79,4 @@
 {"tstamp": 1722433984.3307, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "e93de447a5a24fa494ff1d3829debaa8", "0_model_name": "BM25", "0_prompt": "What is the first probe we sent to space?", "0_output": [["What is the first probe we sent to space?", "Pioneer program\n\nThe Pioneer programs were two series of United States lunar and planetary space probes exploration. The first program, which ran from 1958 to 1960, unsuccessfully attempted to send spacecraft to orbit the Moon, successfully sent one spacecraft to fly by the Moon, and successfully sent one spacecraft to investigate interplanetary space between the orbits of Earth and Venus. The second program, which ran from 1965 to 1992, sent four spacecraft to measure interplanetary space weather, two to explore Jupiter and Saturn, and two to explore Venus. The two outer planet probes, Pioneer 10 and Pioneer 11, became the first two of five artificial objects to achieve the escape velocity that will allow them to leave the Solar System, and carried a golden plaque each depicting a man and a woman and information about the origin and the creators of the probes, in case any extraterrestrials find them someday.\nNaming\nCredit for naming the first probe has been attributed to Stephen A. Saliga, who had been assigned to the Air Force Orientation Group, Wright-Patterson AFB, as chief designer of Air Force exhibits. While he was at a briefing, the spacecraft was described to him, as, a \"lunar-orbiting vehicle, with an infrared scanning device.\" Saliga thought the title too long, and lacked theme for an exhibit design. He suggested, \"Pioneer\", as the name of the probe, since \"the Army had already launched and orbited the Explorer satellite, and their Public Information Office was identifying the Army, as, 'Pioneers in Space,'\" and, by adopting the name, the Air Force would \"make a 'quantum jump' as to who, really, [were] the 'Pioneers' in space.'\""]], "0_corpus": "wikipedia", "1_conv_id": "53b85fd89d2b4f02b5d77aedc69d0628", "1_model_name": "intfloat/e5-mistral-7b-instruct", "1_prompt": "What is the first probe we sent to space?", "1_output": [["What is the first probe we sent to space?", "Sputnik 1\n\nSputnik 1 (, , Satellite 1) was the first artificial Earth satellite. It was launched into an elliptical low Earth orbit by the Soviet Union on 4 October 1957 as part of the Soviet space program. It sent a radio signal back to Earth for three weeks before its three silver-zinc batteries became depleted. Aerodynamic drag caused it to fall back into the atmosphere on 4 January 1958. The world's first observation was made at the school observatory in Rodewisch (Saxony).\nIt was a polished metal sphere  in diameter with four external radio antennas to broadcast radio pulses. Its radio signal was easily detectable by amateur radio operators, and the 65\u00b0 orbital inclination made its flight path cover virtually the entire inhabited Earth."]], "1_corpus": "wikipedia"}
 {"tstamp": 1722435682.4732, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "e3e0803a5ed3495e879fbdcf95b60e6b", "0_model_name": "jinaai/jina-embeddings-v2-base-en", "0_prompt": "Which test was devised to determine whether robots can think?", "0_output": [["Which test was devised to determine whether robots can think?", "Cognitive test\n\nCognitive tests are assessments of the cognitive capabilities of humans and other animals. Tests administered to humans include various forms of IQ tests; those administered to animals include the mirror test (a test of visual self-awareness) and the T maze test (which tests learning ability). Such testing is used in psychology and psychometrics, as well as other fields studying human and animal intelligence."]], "0_corpus": "wikipedia", "1_conv_id": "6b0f0b1750eb4bb5b4313281fda7164d", "1_model_name": "voyage-multilingual-2", "1_prompt": "Which test was devised to determine whether robots can think?", "1_output": [["Which test was devised to determine whether robots can think?", "Turing test\n\nThe test was introduced by Turing in his 1950 paper \"Computing Machinery and Intelligence\" while working at the University of Manchester. It opens with the words: \"I propose to consider the question, 'Can machines think? Because \"thinking\" is difficult to define, Turing chooses to \"replace the question by another, which is closely related to it and is expressed in relatively unambiguous words.\" Turing describes the new form of the problem in terms of a three-person game called the \"imitation game\", in which an interrogator asks questions of a man and a woman in another room in order to determine the correct sex of the two players. Turing's new question is: \"Are there imaginable digital computers which would do well in the imitation game?\" This question, Turing believed, was one that could actually be answered. In the remainder of the paper, he argued against all the major objections to the proposition that \"machines can think\".\nSince Turing introduced his test, it has been both highly influential and widely criticized, and has become an important concept in the philosophy of artificial intelligence. Philosopher John Searle would comment on the Turing test in his Chinese room argument, a thought experiment that stipulates that a machine cannot have a \"mind\", \"understanding\", or \"consciousness\", regardless of how intelligently or human-like the program may make the computer behave. Searle criticizes Turing's test and claims it is insufficient to detect the presence of consciousness.\nHistory"]], "1_corpus": "wikipedia"}
 {"tstamp": 1722435844.6878, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "02c643933a4f421fbf9566730bc39465", "0_model_name": "voyage-multilingual-2", "0_prompt": "Orca life expectancy", "0_output": [["Orca life expectancy", "Orca\n\nLife cycle Female orcas begin to mature at around the age of 10 and reach peak fertility around 20, experiencing periods of polyestrous cycling separated by non-cycling periods of three to 16 months. Females can often breed until age 40, followed by a rapid decrease in fertility. Orcas are among the few animals that undergo menopause and live for decades after they have finished breeding. The lifespans of wild females average 50 to 80 years. Some are claimed to have lived substantially longer: Granny (J2) was estimated by some researchers to have been as old as 105 years at the time of her death, though a biopsy sample indicated her age as 65 to 80 years. It is thought that orcas held in captivity tend to have shorter lives than those in the wild, although this is subject to scientific debate.\nMales mate with females from other pods, which prevents inbreeding. Gestation varies from 15 to 18 months. Mothers usually calve a single offspring about once every five years. In resident pods, births occur at any time of year, although winter is the most common. Mortality is extremely high during the first seven months of life, when 37–50% of all calves die. Weaning begins at about 12 months of age, and is complete by two years. According to observations in several regions, all male and female pod members participate in the care of the young."]], "0_corpus": "wikipedia", "1_conv_id": "b3487cab0720415f8e58696097136d9c", "1_model_name": "embed-english-v3.0", "1_prompt": "Orca life expectancy", "1_output": [["Orca life expectancy", "Captive orcas\n\nDisease and lifespan The lifespan of killer whales in captivity versus wild killer whales is disputed.  Several studies published in scientific journals show that the average mortality rate for captive killer whales is approximately three times higher than in the wild. A 2015 study in the Journal of Mammalogy, authored by SeaWorld's vice-president of theriogenology, Todd Robeck, concluded that the life expectancy for killer whales born at SeaWorld is the same as those in the wild. In the wild, female killer whales have a typical lifespan of 60–80 years, and a maximum recorded lifespan of 103 years.  The average lifespan for males in the wild is 30 years, but some live up to 50–60 years. The 2015 study has been criticised by Trevor Willis, senior lecturer in marine biology at the University of Portsmouth, who stated that the study is misleading, \"clearly wrong\" and indicative of \"poor practice\".  He stated that it is misleading in two ways: \"First, it compares two completely different circumstances: the controlled environment of a swimming pool, with highly trained vets on hand; and the wild ocean. \"There are no predators in a swimming pool. Second, and in the absence of any other information, it appears they've looked at the survival rate of calves in the first two years of life and extrapolated it out 50 years into the future.\" He also stated that no captive orca has lived for 55.8 years, the recorded average life expectancy of adult orcas at SeaWorld."]], "1_corpus": "wikipedia"}
+{"tstamp": 1722435987.9464, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "bc9d5486eaed40f8be91f71650db13c7", "0_model_name": "BM25", "0_prompt": "The best Sentinel-2 bands for wildfire-burned area segmentation", "0_output": [["The best Sentinel-2 bands for wildfire-burned area segmentation", "Title: Sen2Fire: A Challenging Benchmark Dataset for Wildfire Detection using Sentinel Data\n\nAbstract: Utilizing satellite imagery for wildfire detection presents substantial potential for practical applications. To advance the development of machine learning algorithms in this domain, our study introduces the \\textit{Sen2Fire} dataset--a challenging satellite remote sensing dataset tailored for wildfire detection. This dataset is curated from Sentinel-2 multi-spectral data and Sentinel-5P aerosol product, comprising a total of 2466 image patches. Each patch has a size of 512$\\times$512 pixels with 13 bands. Given the distinctive sensitivities of various wavebands to wildfire responses, our research focuses on optimizing wildfire detection by evaluating different wavebands and employing a combination of spectral indices, such as normalized burn ratio (NBR) and normalized difference vegetation index (NDVI). The results suggest that, in contrast to using all bands for wildfire detection, selecting specific band combinations yields superior performance. Additionally, our study underscores the positive impact of integrating Sentinel-5 aerosol data for wildfire detection. The code and dataset are available online (https://zenodo.org/records/10881058)."]], "0_corpus": "arxiv", "1_conv_id": "d679afd9ac124e39a9e06dbe8710a5ff", "1_model_name": "mixedbread-ai/mxbai-embed-large-v1", "1_prompt": "The best Sentinel-2 bands for wildfire-burned area segmentation", "1_output": [["The best Sentinel-2 bands for wildfire-burned area segmentation", "Title: Robust Burned Area Delineation through Multitask Learning\n\nAbstract: In recent years, wildfires have posed a significant challenge due to their increasing frequency and severity. For this reason, accurate delineation of burned areas is crucial for environmental monitoring and post-fire assessment. However, traditional approaches relying on binary segmentation models often struggle to achieve robust and accurate results, especially when trained from scratch, due to limited resources and the inherent imbalance of this segmentation task. We propose to address these limitations in two ways: first, we construct an ad-hoc dataset to cope with the limited resources, combining information from Sentinel-2 feeds with Copernicus activations and other data sources. In this dataset, we provide annotations for multiple tasks, including burned area delineation and land cover segmentation. Second, we propose a multitask learning framework that incorporates land cover classification as an auxiliary task to enhance the robustness and performance of the burned area segmentation models. We compare the performance of different models, including UPerNet and SegFormer, demonstrating the effectiveness of our approach in comparison to standard binary segmentation."]], "1_corpus": "arxiv"}

data/retrieval_individual-f62bcad9-19c4-430b-a7ba-aa2bfeba99dc.jsonl

@@ -328,3 +328,5 @@
 {"tstamp": 1722435799.3056, "task_type": "retrieval", "type": "chat", "model": "BAAI/bge-large-en-v1.5", "gen_params": {}, "start": 1722435799.1268, "finish": 1722435799.3056, "ip": "", "conv_id": "a102e3de078441879dded28975aafc67", "model_name": "BAAI/bge-large-en-v1.5", "prompt": "What is the capital of Australia?", "output": [["What is the capital of Australia?", "Canberra\n\nCanberra ( ) is the capital city of Australia. Founded following the federation of the colonies of Australia as the seat of government for the new nation, it is Australia's largest inland city and the eighth-largest Australian city overall. The city is located at the northern end of the Australian Capital Territory at the northern tip of the Australian Alps, the country's highest mountain range.  Canberra's estimated population was 466,566.\nThe area chosen for the capital had been inhabited by Aboriginal Australians for up to 21,000 years, by groups including the Ngunnawal, Ngunawal and Ngambri. European settlement commenced in the first half of the 19th century, as evidenced by surviving landmarks such as St John's Anglican Church and Blundells Cottage. On 1 January 1901, federation of the colonies of Australia was achieved. Following a long dispute over whether Sydney or Melbourne should be the national capital, a compromise was reached: the new capital would be built in New South Wales, so long as it was at least  from Sydney. The capital city was founded and formally named as Canberra in 1913. A plan by the American architects Walter Burley Griffin and Marion Mahony Griffin was selected after an international design contest, and construction commenced in 1913. Unusual among Australian cities, it is an entirely planned city. The Griffins' plan featured geometric motifs and was centred on axes aligned with significant topographical landmarks such as Black Mountain, Mount Ainslie, Capital Hill and City Hill. Canberra's mountainous location makes it the only mainland Australian city where snow-capped mountains can be seen in winter, although snow in the city itself is uncommon."]], "corpus": "wikipedia"}
 {"tstamp": 1722435824.195, "task_type": "retrieval", "type": "chat", "model": "voyage-multilingual-2", "gen_params": {}, "start": 1722435823.6918, "finish": 1722435824.195, "ip": "", "conv_id": "02c643933a4f421fbf9566730bc39465", "model_name": "voyage-multilingual-2", "prompt": "Orca life expectancy", "output": [["Orca life expectancy", "Orca\n\nLife cycle Female orcas begin to mature at around the age of 10 and reach peak fertility around 20, experiencing periods of polyestrous cycling separated by non-cycling periods of three to 16 months. Females can often breed until age 40, followed by a rapid decrease in fertility. Orcas are among the few animals that undergo menopause and live for decades after they have finished breeding. The lifespans of wild females average 50 to 80 years. Some are claimed to have lived substantially longer: Granny (J2) was estimated by some researchers to have been as old as 105 years at the time of her death, though a biopsy sample indicated her age as 65 to 80 years. It is thought that orcas held in captivity tend to have shorter lives than those in the wild, although this is subject to scientific debate.\nMales mate with females from other pods, which prevents inbreeding. Gestation varies from 15 to 18 months. Mothers usually calve a single offspring about once every five years. In resident pods, births occur at any time of year, although winter is the most common. Mortality is extremely high during the first seven months of life, when 37–50% of all calves die. Weaning begins at about 12 months of age, and is complete by two years. According to observations in several regions, all male and female pod members participate in the care of the young."]], "corpus": "wikipedia"}
 {"tstamp": 1722435824.195, "task_type": "retrieval", "type": "chat", "model": "embed-english-v3.0", "gen_params": {}, "start": 1722435823.6918, "finish": 1722435824.195, "ip": "", "conv_id": "b3487cab0720415f8e58696097136d9c", "model_name": "embed-english-v3.0", "prompt": "Orca life expectancy", "output": [["Orca life expectancy", "Captive orcas\n\nDisease and lifespan The lifespan of killer whales in captivity versus wild killer whales is disputed.  Several studies published in scientific journals show that the average mortality rate for captive killer whales is approximately three times higher than in the wild. A 2015 study in the Journal of Mammalogy, authored by SeaWorld's vice-president of theriogenology, Todd Robeck, concluded that the life expectancy for killer whales born at SeaWorld is the same as those in the wild. In the wild, female killer whales have a typical lifespan of 60–80 years, and a maximum recorded lifespan of 103 years.  The average lifespan for males in the wild is 30 years, but some live up to 50–60 years. The 2015 study has been criticised by Trevor Willis, senior lecturer in marine biology at the University of Portsmouth, who stated that the study is misleading, \"clearly wrong\" and indicative of \"poor practice\".  He stated that it is misleading in two ways: \"First, it compares two completely different circumstances: the controlled environment of a swimming pool, with highly trained vets on hand; and the wild ocean. \"There are no predators in a swimming pool. Second, and in the absence of any other information, it appears they've looked at the survival rate of calves in the first two years of life and extrapolated it out 50 years into the future.\" He also stated that no captive orca has lived for 55.8 years, the recorded average life expectancy of adult orcas at SeaWorld."]], "corpus": "wikipedia"}
+{"tstamp": 1722435933.2884, "task_type": "retrieval", "type": "chat", "model": "BM25", "gen_params": {}, "start": 1722435933.1316, "finish": 1722435933.2884, "ip": "", "conv_id": "bc9d5486eaed40f8be91f71650db13c7", "model_name": "BM25", "prompt": "The best Sentinel-2 bands for wildfire-burned area segmentation", "output": [["The best Sentinel-2 bands for wildfire-burned area segmentation", "Title: Sen2Fire: A Challenging Benchmark Dataset for Wildfire Detection using Sentinel Data\n\nAbstract: Utilizing satellite imagery for wildfire detection presents substantial potential for practical applications. To advance the development of machine learning algorithms in this domain, our study introduces the \\textit{Sen2Fire} dataset--a challenging satellite remote sensing dataset tailored for wildfire detection. This dataset is curated from Sentinel-2 multi-spectral data and Sentinel-5P aerosol product, comprising a total of 2466 image patches. Each patch has a size of 512$\\times$512 pixels with 13 bands. Given the distinctive sensitivities of various wavebands to wildfire responses, our research focuses on optimizing wildfire detection by evaluating different wavebands and employing a combination of spectral indices, such as normalized burn ratio (NBR) and normalized difference vegetation index (NDVI). The results suggest that, in contrast to using all bands for wildfire detection, selecting specific band combinations yields superior performance. Additionally, our study underscores the positive impact of integrating Sentinel-5 aerosol data for wildfire detection. The code and dataset are available online (https://zenodo.org/records/10881058)."]], "corpus": "arxiv"}
+{"tstamp": 1722435933.2884, "task_type": "retrieval", "type": "chat", "model": "mixedbread-ai/mxbai-embed-large-v1", "gen_params": {}, "start": 1722435933.1316, "finish": 1722435933.2884, "ip": "", "conv_id": "d679afd9ac124e39a9e06dbe8710a5ff", "model_name": "mixedbread-ai/mxbai-embed-large-v1", "prompt": "The best Sentinel-2 bands for wildfire-burned area segmentation", "output": [["The best Sentinel-2 bands for wildfire-burned area segmentation", "Title: Robust Burned Area Delineation through Multitask Learning\n\nAbstract: In recent years, wildfires have posed a significant challenge due to their increasing frequency and severity. For this reason, accurate delineation of burned areas is crucial for environmental monitoring and post-fire assessment. However, traditional approaches relying on binary segmentation models often struggle to achieve robust and accurate results, especially when trained from scratch, due to limited resources and the inherent imbalance of this segmentation task. We propose to address these limitations in two ways: first, we construct an ad-hoc dataset to cope with the limited resources, combining information from Sentinel-2 feeds with Copernicus activations and other data sources. In this dataset, we provide annotations for multiple tasks, including burned area delineation and land cover segmentation. Second, we propose a multitask learning framework that incorporates land cover classification as an auxiliary task to enhance the robustness and performance of the burned area segmentation models. We compare the performance of different models, including UPerNet and SegFormer, demonstrating the effectiveness of our approach in comparison to standard binary segmentation."]], "corpus": "arxiv"}