76 research outputs found
For What It's Worth: Humans Overwrite Their Economic Self-interest to Avoid Bargaining With AI Systems
No full textAs algorithms are increasingly augmenting and substituting human decision-making, understanding how the introduction of computational agents changes the fundamentals of human behavior becomes vital. This pertains to not only users, but also those parties who face the consequences of an algorithmic decision. In a controlled experiment with 480 participants, we exploit an extended version of two-player ultimatum bargaining where responders choose to bargain with either another human, another human with an AI decision aid or an autonomous AI-system acting on behalf of a passive human proposer. Our results show strong responder preferences against the algorithm, as most responders opt for a human opponent and demand higher compensation to reach a contract with autonomous agents. To map these preferences to economic expectations, we elicit incentivized subject beliefs about their opponent's behavior. The majority of responders maximize their expected value when this is line with approaching the human proposer. In contrast, responders predicting income maximization for the autonomous AI-system overwhelmingly override economic self-interest to avoid the algorithm.Web Information System
From novella to theatre and opera: translating 'otherness' in Cavalleria rusticana
No full textThis chapter examines what it meant to perform translation in Cavalleria rusticana. It examines: 1) the original novella written by Giovanni Verga in 1880; 2) the 1884 stage version adapted by the author and interpreted by a variety of star actors, including Eleonora Duse and the Sicilian dialect players; and 3) the 1890 operatic version composed by Pietro Mascagni and performed by celebrity sopranos, such as Emma Calvé. Through close examination of newspaper reviews and early accounts, it is argued that performing translation meant generating and circulating an exoticized ‘brand’ of Sicilian-ness in and outside the Italian peninsula shortly after political unification in 1861. This chapter thus offers new perspectives into questions of racial stereotypes, and a provides basis for new insights into the Sicilian dialect players, in particular, who, pioneering a physical and bodily form of communication, transcended language barriers and mediated the foreign text through a kind of translation that went beyond the written word
The Titans at SemEval-2019 Task 5: Detection of hate speech against immigrants and women in Twitter
No full textThe Titans at SemEval-2019 Task 6: Offensive Language Identification, Categorization and Target Identification
No full textCheirostylis parvifolia Lindl.
No full textCheirostylis parvifolia Lindl.inEdwards ’s Bot.Reg. 25: 19. 1839. Typus: SRI LANKA “ Ceylon ”: Loddiges s.n. (holo-: K [K000718267], photo!). = Cheirostylis seidenfadeniana C. S. Kumar & F. N. Rasm. in Nordic J. Bot. 7: 409. 1987. Typus: INDIA. Kerala: Ponmudi, Trivandrum Dist., 950 m, 25.X. 1983, C. Sathish Kumar CU 36960 (holo-: TBGT; iso-: C, CALI), syn. nov.. Distribution. – India: Kerala, Maharashtra (fide JAYAWEERA, 1981;P UNEKAR, 2002);SriLanka. Note. – SATHISH KUMAR & RASMUSSEN (1987) distinguished C. seidenfadeniana from C. parvifolia on the basis of epichile with entire lobules and 2 conspicuous tufts of hairs at its base. The authors mentioned that the material of C. parvifolia was very scarce from India and they could consult only 3 old herbarium specimens including the type (Loddiges s.n., K) which was in very bad condition. However, it has been found that the holotype including its associated illustrations of flower and dissected floral parts (on the type sheet) is sufficient for its purpose and there is no need of epitypification at present. During the present work, critical study of the live specimens of C. seidenfadeniana provided by Dr. C. Sathish Kumar (first author of C. seidenfadeniana) revealed that the minute tufts of hairs at the base of epichile can be seen only in live, pickled or very well preserved specimens but not in the old herbarium specimens. The present study also reveals that the margin of the epichile lobules of C. seidenfadeniana varies (Fig. 1) from nearly entire to 2-4 lacerate which clearly indicates that C. seidenfadeniana and C. parvifolia are the same species. Thus, C. seidenfadeniana is treated here as synonym of C. parvifolia. In spite of author’s personal visit to CALI and TBGT, the holotype and isotype(s) of C. seidenfadeniana could not be located as claimed in the protologue. Dr. Olof Ryding, Curator of Vascular Plants of C also confirmed the unavailability of any type specimen of C. seidenfadeniana in C. Surprisingly, a specimen bearing the same collection number and field data (excepting the date which is prior to publication of C. seidenfadeniana) as that of the holotype and isotypes of C. seidenfadeniana has been located at K. In absence of all the designated types (holotype and isotypes), the K-specimen may be selected as lectotype (if there is no doubt that it was actually used while describing C. seidenfadeniana) or as neotype of C. seidenfadeniana. For the time being the lecto-or neotypification has not been done in the present paper with a hope that the types may be found in near future. This rare species is presently grown and conserved in the Orchidarium of Tropical Botanic Garden and Research Institute, Thiruvananthapuram, Kerala, India by Dr. Sathish Kumar and his co-workers.Published as part of Bhattacharjee, Avishek, 2012, On the status of some species of Cheirostylis Blume (Orchidaceae) from India, pp. 31-35 in Candollea 67 (1) on page 32, DOI: 10.15553/c2012v671a3, http://zenodo.org/record/570865
A Comparison of Instance Attribution Methods: Comparing Instance Attribution Methods to Baseline k-Nearest Neighbors Method
No full textIn this research, a comparison between different Instance Attribution (IA) methods and k-Nearest Neighbors (kNN) via cosine similarity is conducted on a Natural Language Processing (NLP) machine learning model. The format in which the comparison is made is by way of a human survey and automated similarity comparisons of representative vectors. The goal of this is to judge and compare the effectiveness of each method’s results in the context of a human’s language understanding and ability to determine if a fact is true or not. Through this research, it was found that for results obtained on the same input, IA methods were preferred 32.5% more often than kNN. It is also shown that this preference is not linked to the similarity between the IA results and the kNN results. Through these findings, it can be seen that when understood through the lens of human comprehension, IA methods are much more effective at generating a set of influential training points from the model’s training dataset.CSE3000 Research ProjectComputer Science and Engineerin
Explore-then-Commit Algorithms for Decentralized Two-Sided Matching Markets
No full textOnline learning in a decentralized two-sided matching markets, where the demand-side (players) compete to match with the supply-side (arms), has received substantial interest because it abstracts out the complex interactions in matching platforms (e.g. UpWork, TaskRabbit). However, past works assume that each arm knows their preference ranking over the players (one-sided learning), and each player aim to learn the preference over arms through successive interactions. Moreover, several (impractical) assumptions on the problem are usually made for theoretical tractability such as broadcast player-arm match Liu et al. (2020; 2021); Kong & Li (2023) or serial dictatorship Sankararaman et al. (2021); Basu et al. (2021); Ghosh et al. (2022). In this paper, we study a decentralized two-sided matching market, where we do not assume that the preference ranking over players are known to the arms apriori. Furthermore, we do not have any structural assumptions on the problem. We propose a multi-phase explore-then-commit type algorithm namely epoch-based CA-ETC (collision avoidance explore then commit) (\texttt{CA-ETC} in short) for this problem that does not require any communication across agents (players and arms) and hence decentralized. We show that for the initial epoch length of and subsequent epoch-lengths of (for the th epoch with as an input parameter to the algorithm), \texttt{CA-ETC} yields a player optimal expected regret of for the -th player, where is the learning horizon, is the number of arms and is an appropriately defined problem gap. Furthermore, we propose a blackboard communication based baseline achieving logarithmic regret in .Accepted at International Symposium of Information Theory (ISIT) 202
Are BERT-based fact-checking models robust against adversarial attack?
No full textWe seek to examine the vulnerability of BERT-based fact-checking. We implement a gradient based, adversarial attack strategy, based on Hotflip swapping individual tokens from the input. We use this on a pre-trained ExPred model for fact-checking. We find that gradient based adversarial attacks are ineffective against ExPred. Uncertainties about the similitude of the examples generated by our adversarial attack implementation cast doubts on the results.https://github.com/somePersone/HotFlip-for-Expred Hotflip implementationCSE3000 Research ProjectComputer Science and Engineerin
Gravitational pulsars: correlations between the electromagnetic and the continuous gravitational wave signal
No full textInternational audienceNeutron stars emitting continuous gravitational waves may be regarded as gravitational pulsars, in the sense that it could be possible to track the evolution of their rotational period with long-baseline observations of next-generation gravitational wave interferometers. Assuming that the pulsar's electromagnetic signal is tracked and allows us to monitor the pulsar's spin evolution, we provide a physical interpretation of the possible observed correlation between this timing solution and its gravitational counterpart, if the system is also detected in gravitational waves. In particular, we show that next-generation detectors, such as the Einstein Telescope, could have the sensitivity to discern different models for the coupling between the superfluid and normal components of the neutron star and constrain the origin of timing noise (whether due to magnetospheric or internal processes). Observational confirmation of one of the proposed scenarios would therefore provide valuable information on the physics of gravitational wave emission from pulsars
- …
