Scheduled Commit

data/retrieval_battle-407c4836-37e2-4f9f-8e9b-06706cc4440c.jsonl

@@ -2,3 +2,4 @@
 {"tstamp": 1722441638.7331, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "03187db5197c4fe7b645e21973a23910", "0_model_name": "mixedbread-ai/mxbai-embed-large-v1", "0_prompt": "can minecraft wii u play with xbox?", "0_output": [["can minecraft wii u play with xbox?", "Gaming Stackexchange\n\nQ: Can you play minecraft on Wii U and Xbox together on mulitplayer? I want to play with my friend but don't know how!\n\nA: What you are referring to is called \"Cross Platform Multiplayer\" Where two or more players on different consoles can play the same game. Currently, Cross platform multiplayer is very rare, and few games have this feature available.\nThere have been rumors that improved Cross platform multiplayer may be included at a future date on some or all of the minecraft versions, but this is not yet confirmed.(http://www.ign.com/articles/2015/08/05/gamescom-2015-mojang-wants-minecraft-to-work-cross-platform) \nUnfortunatly, at the moment, You can only play Minecraft online multiplayer if you, and everyone you are trying to play with are using the same version of minecraft on the same console or platform. \n*EDIT\nAs Pointed out by colorfusion in the comments, Minecraft Pocket Edition is compatible with Minecraft Windows 10 Edition. (Here's an article about it http://www.pcworld.com/article/2982932/software-games/minecraft-pocket-edition-now-lets-windows-10-ios-and-android-gamers-play-together.html)\n"]], "0_corpus": "stackexchange", "1_conv_id": "4405fb6401d8453189dbfe07d1e675e3", "1_model_name": "GritLM/GritLM-7B", "1_prompt": "can minecraft wii u play with xbox?", "1_output": [["can minecraft wii u play with xbox?", "Gaming Stackexchange\n\nQ: Can you play minecraft on Wii U and Xbox together on mulitplayer? I want to play with my friend but don't know how!\n\nA: What you are referring to is called \"Cross Platform Multiplayer\" Where two or more players on different consoles can play the same game. Currently, Cross platform multiplayer is very rare, and few games have this feature available.\nThere have been rumors that improved Cross platform multiplayer may be included at a future date on some or all of the minecraft versions, but this is not yet confirmed.(http://www.ign.com/articles/2015/08/05/gamescom-2015-mojang-wants-minecraft-to-work-cross-platform) \nUnfortunatly, at the moment, You can only play Minecraft online multiplayer if you, and everyone you are trying to play with are using the same version of minecraft on the same console or platform. \n*EDIT\nAs Pointed out by colorfusion in the comments, Minecraft Pocket Edition is compatible with Minecraft Windows 10 Edition. (Here's an article about it http://www.pcworld.com/article/2982932/software-games/minecraft-pocket-edition-now-lets-windows-10-ios-and-android-gamers-play-together.html)\n"]], "1_corpus": "stackexchange"}
 {"tstamp": 1722441688.0512, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "6b45321f7e8743f6ba37ad8fbb58bcba", "0_model_name": "embed-english-v3.0", "0_prompt": "A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "0_output": [["A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "Title: Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue\n\nAbstract: Knowledge-grounded dialogue is a task of generating an informative response based on both discourse context and external knowledge. As we focus on better modeling the knowledge selection in the multi-turn knowledge-grounded dialogue, we propose a sequential latent variable model as the first approach to this matter. The model named sequential knowledge transformer (SKT) can keep track of the prior and posterior distribution over knowledge; as a result, it can not only reduce the ambiguity caused from the diversity in knowledge selection of conversation but also better leverage the response information for proper choice of knowledge. Our experimental results show that the proposed model improves the knowledge selection accuracy and subsequently the performance of utterance generation. We achieve the new state-of-the-art performance on Wizard of Wikipedia (Dinan et al., 2019) as one of the most large-scale and challenging benchmarks. We further validate the effectiveness of our model over existing conversation methods in another knowledge-based dialogue Holl-E dataset (Moghe et al., 2018)."]], "0_corpus": "arxiv", "1_conv_id": "d4ac56d1a9f84dc5ba91f489a57a6eac", "1_model_name": "GritLM/GritLM-7B", "1_prompt": "A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "1_output": [["A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "Title: Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue\n\nAbstract: Knowledge-grounded dialogue is a task of generating an informative response based on both discourse context and external knowledge. As we focus on better modeling the knowledge selection in the multi-turn knowledge-grounded dialogue, we propose a sequential latent variable model as the first approach to this matter. The model named sequential knowledge transformer (SKT) can keep track of the prior and posterior distribution over knowledge; as a result, it can not only reduce the ambiguity caused from the diversity in knowledge selection of conversation but also better leverage the response information for proper choice of knowledge. Our experimental results show that the proposed model improves the knowledge selection accuracy and subsequently the performance of utterance generation. We achieve the new state-of-the-art performance on Wizard of Wikipedia (Dinan et al., 2019) as one of the most large-scale and challenging benchmarks. We further validate the effectiveness of our model over existing conversation methods in another knowledge-based dialogue Holl-E dataset (Moghe et al., 2018)."]], "1_corpus": "arxiv"}
 {"tstamp": 1722441726.5986, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "0eb1b97ea1ea4813bcb0d07561d9762b", "0_model_name": "jinaai/jina-embeddings-v2-base-en", "0_prompt": "Paper exploring algorithmic fidelity in language models for social science research.", "0_output": [["Paper exploring algorithmic fidelity in language models for social science research.", "Title: Analysis and modeling of scale-invariance in plankton abundance\n\nAbstract: The power spectrum, $S$, of horizontal transects of plankton abundance are often observed to have a power-law dependence on wavenumber, $k$, with exponent close to -2: $S(k)\\propto k^{-2}$ over a wide range of scales. I present power spectral analyses of aircraft lidar measurements of phytoplankton abundance from scales of 1 to 100 km. A power spectrum $S(k)\\propto k^{-2}$ is obtained. As a model for this observation, I consider a stochastic growth equation where the rate of change of plankton abundance is determined by turbulent mixing, modeled as a diffusion process in two dimensions, and exponential growth with a stochastically variable net growth rate representing a fluctuating environment. The model predicts a lognormal distribution of abundance and a power spectrum of horizontal transects $S(k)\\propto k^{-1.8}$, close to the observed spectrum. The model equation predicts that the power spectrum of variations in abundance in time at a point in space is $S(f)\\propto f^{-1.5}$ (where $f$ is the frequency). Time series analysis of local variations of phytoplankton and zooplankton yield a power-law power spectrum with exponents -1.3 and -1.2, respectively from time scales of one hour to one year. These values are roughly consistent with the model prediction of -1.5. The distribution of abundances is nearly lognormal as predicted. The model may be more generally applicable than for the spatial distribution of plankton. I relate the model predictions to observations of spatial patchiness in vegetation."]], "0_corpus": "arxiv", "1_conv_id": "519f89cd54734d5a8b84fc365f499beb", "1_model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "1_prompt": "Paper exploring algorithmic fidelity in language models for social science research.", "1_output": [["Paper exploring algorithmic fidelity in language models for social science research.", "Title: Neural language representations predict outcomes of scientific research\n\nAbstract: Many research fields codify their findings in standard formats, often by reporting correlations between quantities of interest. But the space of all testable correlates is far larger than scientific resources can currently address, so the ability to accurately predict correlations would be useful to plan research and allocate resources. Using a dataset of approximately 170,000 correlational findings extracted from leading social science journals, we show that a trained neural network can accurately predict the reported correlations using only the text descriptions of the correlates. Accurate predictive models such as these can guide scientists towards promising untested correlates, better quantify the information gained from new findings, and has implications for moving artificial intelligence systems from predicting structures to predicting relationships in the real world."]], "1_corpus": "arxiv"}
+{"tstamp": 1722442171.6617, "task_type": "retrieval", "type": "bothbadvote", "models": ["", ""], "ip": "", "0_conv_id": "60025466ec7e44b1b64b3e8cf40ca93a", "0_model_name": "BM25", "0_prompt": "Har Mette Frederiksen søskende?", "0_output": [["Har Mette Frederiksen søskende?", "Mette Frederiksen\n\n2019 election The 2019 general election saw the Social Democrats gaining a further seat while support for the Danish People's Party and the Liberal Alliance collapsed, costing Lars Løkke Rasmussen his majority. With the result beyond doubt on election night, Rasmussen conceded defeat. Frederiksen was appointed prime minister on 27 June 2019, heading an exclusively Social Democratic minority government supported by the red bloc of the Social Liberal Party, the Red-Green Alliance and the Green Left. Despite having run on an anti-immigration stance during the election, Frederiksen briefly shifted her stance on immigration by allowing more foreign labour and reversing government plans to hold foreign criminals offshore after winning government.\n2020 mink cull\nAt a press conference on 4 November 2020, Mette Frederiksen stated that the government had decided that all mink in Denmark should be killed due to the risk of infection with COVID-19. Subsequently, it emerged that this order was illegal and was described by many as being in violation of the Danish Constitution. The government came up with changing explanations and several parties in the Folketing demanded an account of the mink case. The statement was published on 18 November 2020 and it emerged that six ministers had been warned on 1 October 2020 that the order was illegal. Minister of Food, Agriculture and Fisheries Mogens Jensen withdrew immediately. Mette Frederiksen has denied knowledge of the lack of legal basis."]], "0_corpus": "wikipedia", "1_conv_id": "06d026b5e1be4d099e2309f63923c321", "1_model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "1_prompt": "Har Mette Frederiksen søskende?", "1_output": [["Har Mette Frederiksen søskende?", "Hans Christian Gram\n\nGram stain In Berlin, in 1884, Gram developed a method for distinguishing between two major classes of bacteria. This technique, known as Gram staining, continues to be a standard procedure of medical microbiology. This work gained Gram an international reputation. The staining method later played a major role in classifying bacteria. Gram was a modest man, and in his initial publication he remarked, \"I have therefore published the method, although I am aware that as yet it is very defective and imperfect; but it is hoped that also in the hands of other investigators it will turn out to be useful.\"\nA Gram stain is made using a primary stain of crystal violet and a counterstain of safranin. Bacteria that turn purple when stained are termed 'Gram-positive', while those that turn red when counterstained are termed 'Gram-negative'.\nOther work\nGram's initial work concerned the study of human red blood cells. He was among the first to recognise that macrocytes were characteristic of pernicious anaemia.\nDuring 1891, Gram taught pharmacology, and later that year was appointed professor at the University of Copenhagen. In 1900, he resigned his professorship of pharmacology to become professor of medicine. As a professor, he published four volumes of clinical lectures which became used widely in Denmark. He retired from the University of Copenhagen in 1923, and died in 1938.\nPopular recognition\nOn 13 September 2019, Google commemorated the anniversary of his birth with a Doodle for Canada, Peru, Argentina, Australia, New Zealand, Israel, India and some European countries.\nPersonal life\nGram's great-granddaughter Lone Gram is a microbiologist who works at the Technical University of Denmark."]], "1_corpus": "wikipedia"}

data/retrieval_individual-407c4836-37e2-4f9f-8e9b-06706cc4440c.jsonl

@@ -6,3 +6,5 @@
 {"tstamp": 1722441665.4403, "task_type": "retrieval", "type": "chat", "model": "GritLM/GritLM-7B", "gen_params": {}, "start": 1722441647.9506, "finish": 1722441665.4403, "ip": "", "conv_id": "d4ac56d1a9f84dc5ba91f489a57a6eac", "model_name": "GritLM/GritLM-7B", "prompt": "A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "output": [["A novel approach for knowledge-grounded dialogues by improving knowledge selection with a sequential latent variable model.", "Title: Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue\n\nAbstract: Knowledge-grounded dialogue is a task of generating an informative response based on both discourse context and external knowledge. As we focus on better modeling the knowledge selection in the multi-turn knowledge-grounded dialogue, we propose a sequential latent variable model as the first approach to this matter. The model named sequential knowledge transformer (SKT) can keep track of the prior and posterior distribution over knowledge; as a result, it can not only reduce the ambiguity caused from the diversity in knowledge selection of conversation but also better leverage the response information for proper choice of knowledge. Our experimental results show that the proposed model improves the knowledge selection accuracy and subsequently the performance of utterance generation. We achieve the new state-of-the-art performance on Wizard of Wikipedia (Dinan et al., 2019) as one of the most large-scale and challenging benchmarks. We further validate the effectiveness of our model over existing conversation methods in another knowledge-based dialogue Holl-E dataset (Moghe et al., 2018)."]], "corpus": "arxiv"}
 {"tstamp": 1722441702.2496, "task_type": "retrieval", "type": "chat", "model": "jinaai/jina-embeddings-v2-base-en", "gen_params": {}, "start": 1722441701.8684, "finish": 1722441702.2496, "ip": "", "conv_id": "0eb1b97ea1ea4813bcb0d07561d9762b", "model_name": "jinaai/jina-embeddings-v2-base-en", "prompt": "Paper exploring algorithmic fidelity in language models for social science research.", "output": [["Paper exploring algorithmic fidelity in language models for social science research.", "Title: Analysis and modeling of scale-invariance in plankton abundance\n\nAbstract: The power spectrum, $S$, of horizontal transects of plankton abundance are often observed to have a power-law dependence on wavenumber, $k$, with exponent close to -2: $S(k)\\propto k^{-2}$ over a wide range of scales. I present power spectral analyses of aircraft lidar measurements of phytoplankton abundance from scales of 1 to 100 km. A power spectrum $S(k)\\propto k^{-2}$ is obtained. As a model for this observation, I consider a stochastic growth equation where the rate of change of plankton abundance is determined by turbulent mixing, modeled as a diffusion process in two dimensions, and exponential growth with a stochastically variable net growth rate representing a fluctuating environment. The model predicts a lognormal distribution of abundance and a power spectrum of horizontal transects $S(k)\\propto k^{-1.8}$, close to the observed spectrum. The model equation predicts that the power spectrum of variations in abundance in time at a point in space is $S(f)\\propto f^{-1.5}$ (where $f$ is the frequency). Time series analysis of local variations of phytoplankton and zooplankton yield a power-law power spectrum with exponents -1.3 and -1.2, respectively from time scales of one hour to one year. These values are roughly consistent with the model prediction of -1.5. The distribution of abundances is nearly lognormal as predicted. The model may be more generally applicable than for the spatial distribution of plankton. I relate the model predictions to observations of spatial patchiness in vegetation."]], "corpus": "arxiv"}
 {"tstamp": 1722441702.2496, "task_type": "retrieval", "type": "chat", "model": "Alibaba-NLP/gte-Qwen2-7B-instruct", "gen_params": {}, "start": 1722441701.8684, "finish": 1722441702.2496, "ip": "", "conv_id": "519f89cd54734d5a8b84fc365f499beb", "model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "prompt": "Paper exploring algorithmic fidelity in language models for social science research.", "output": [["Paper exploring algorithmic fidelity in language models for social science research.", "Title: Neural language representations predict outcomes of scientific research\n\nAbstract: Many research fields codify their findings in standard formats, often by reporting correlations between quantities of interest. But the space of all testable correlates is far larger than scientific resources can currently address, so the ability to accurately predict correlations would be useful to plan research and allocate resources. Using a dataset of approximately 170,000 correlational findings extracted from leading social science journals, we show that a trained neural network can accurately predict the reported correlations using only the text descriptions of the correlates. Accurate predictive models such as these can guide scientists towards promising untested correlates, better quantify the information gained from new findings, and has implications for moving artificial intelligence systems from predicting structures to predicting relationships in the real world."]], "corpus": "arxiv"}
+{"tstamp": 1722442130.6514, "task_type": "retrieval", "type": "chat", "model": "BM25", "gen_params": {}, "start": 1722442130.3471, "finish": 1722442130.6514, "ip": "", "conv_id": "60025466ec7e44b1b64b3e8cf40ca93a", "model_name": "BM25", "prompt": "Har Mette Frederiksen søskende?", "output": [["Har Mette Frederiksen søskende?", "Mette Frederiksen\n\n2019 election The 2019 general election saw the Social Democrats gaining a further seat while support for the Danish People's Party and the Liberal Alliance collapsed, costing Lars Løkke Rasmussen his majority. With the result beyond doubt on election night, Rasmussen conceded defeat. Frederiksen was appointed prime minister on 27 June 2019, heading an exclusively Social Democratic minority government supported by the red bloc of the Social Liberal Party, the Red-Green Alliance and the Green Left. Despite having run on an anti-immigration stance during the election, Frederiksen briefly shifted her stance on immigration by allowing more foreign labour and reversing government plans to hold foreign criminals offshore after winning government.\n2020 mink cull\nAt a press conference on 4 November 2020, Mette Frederiksen stated that the government had decided that all mink in Denmark should be killed due to the risk of infection with COVID-19. Subsequently, it emerged that this order was illegal and was described by many as being in violation of the Danish Constitution. The government came up with changing explanations and several parties in the Folketing demanded an account of the mink case. The statement was published on 18 November 2020 and it emerged that six ministers had been warned on 1 October 2020 that the order was illegal. Minister of Food, Agriculture and Fisheries Mogens Jensen withdrew immediately. Mette Frederiksen has denied knowledge of the lack of legal basis."]], "corpus": "wikipedia"}
+{"tstamp": 1722442130.6514, "task_type": "retrieval", "type": "chat", "model": "Alibaba-NLP/gte-Qwen2-7B-instruct", "gen_params": {}, "start": 1722442130.3471, "finish": 1722442130.6514, "ip": "", "conv_id": "06d026b5e1be4d099e2309f63923c321", "model_name": "Alibaba-NLP/gte-Qwen2-7B-instruct", "prompt": "Har Mette Frederiksen søskende?", "output": [["Har Mette Frederiksen søskende?", "Hans Christian Gram\n\nGram stain In Berlin, in 1884, Gram developed a method for distinguishing between two major classes of bacteria. This technique, known as Gram staining, continues to be a standard procedure of medical microbiology. This work gained Gram an international reputation. The staining method later played a major role in classifying bacteria. Gram was a modest man, and in his initial publication he remarked, \"I have therefore published the method, although I am aware that as yet it is very defective and imperfect; but it is hoped that also in the hands of other investigators it will turn out to be useful.\"\nA Gram stain is made using a primary stain of crystal violet and a counterstain of safranin. Bacteria that turn purple when stained are termed 'Gram-positive', while those that turn red when counterstained are termed 'Gram-negative'.\nOther work\nGram's initial work concerned the study of human red blood cells. He was among the first to recognise that macrocytes were characteristic of pernicious anaemia.\nDuring 1891, Gram taught pharmacology, and later that year was appointed professor at the University of Copenhagen. In 1900, he resigned his professorship of pharmacology to become professor of medicine. As a professor, he published four volumes of clinical lectures which became used widely in Denmark. He retired from the University of Copenhagen in 1923, and died in 1938.\nPopular recognition\nOn 13 September 2019, Google commemorated the anniversary of his birth with a Doodle for Canada, Peru, Argentina, Australia, New Zealand, Israel, India and some European countries.\nPersonal life\nGram's great-granddaughter Lone Gram is a microbiologist who works at the Technical University of Denmark."]], "corpus": "wikipedia"}