A personal communication

By Micheal M. Jacob. 1989

Subfields: Cognitive Science & Psychology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Science & Psychology

Citations in the corpus

Carbon Emissions and Large Neural Network Training (2021, methodology): “OpenAI measured GPT-3 in a Microsoft Azure datacenter [Sut21].”
RNADE: The real-valued neural autoregressive density-estimator (2013, methodology): “Recent work by Zoran and Weiss [21], projects the data on the leading 63 eigenvectors of each component, when measuring the model likelihood [27].”
Feedforward semantic segmentation with zoom-out features (2014, background): “Recent advances along these lines include [15], which uses convnets and [8], which improves upon the re-ranking in [39], also using convnet-based features.”
Automatic Evaluation of Summaries Using N-gram Co-occurrence Statistics (2003, methodology): “According to Over (2003), NIST spent about 3,000 man hours each in DUC 2001 and 2002 for topic and document selection, summary creation, and manual evaluation.”
The power of synergy in differential privacy: Combining a small curator with local randomizers (2019, methodology): “We will use the following multiplicative variant of their result [52], whose proof is a variant of the original proof of [7].”
Minimum Probability Flow Learning (2009, background): “In addition, K ( θ ) is convex for all models p ( ∞ ) ( θ ) in the exponential family - that is, models whose energy functions E ( θ ) are linear in their parameters θ [13] (see Appendix B).”
Tight bounds for worst-case equilibria (2002, background): “We show that for m identical links the worst-case coordination ratio is at most O ( log m/ log log m ) . signiﬁcantly The same upper bound was presented by Koutsoupias et al. [KMS01] with a longer proof.”
Energy Efficiency in Buildings Through Information - Swedish Perspective (2000, background): “…energy advis-ors, property technicians and chimney sweeps, 3. consulting and contractor authorities and other organisations (NUTEK, 1996a,b; FSE, 1998; Eriksson, The common goal for the future is to ‘ act towards a more e $ cient energy use, in # uence and create conditions for environmentally…”
Representation Mixing for TTS Synthesis (2018, unknown): “This process optimizes randomly initialized parameters (representing the desired wave-form) passed through a log mel spectrogram transform, to make the transform of these parameters match a given log mel spectrogram [26].”
Quantum chaos and the complexity of spread of states (2022, unknown): “Alternatively, we can take suﬃciently many instances of random Hamiltonians with their associated spectra, and then construct another spectrum by randomly sampling eigenvalues from the diﬀerent Hamiltonians [83].”
Microreboot - A Technique for Cheap Recovery (2004, methodology): “Fortunately, it is mostly application-level failures that bring down enterprise-scale software [32, 13, 35, 36], while the underlying platform (hardware and operating system) is reliable, by comparison.”
Valuing ecosystem and economic services across land-use scenarios in the Prairie Pothole Region of the Dakotas, USA (2011, unknown): “Based on conversations with the USDA (i.e. Hyberg, 2009), it is not politically feasible to go much beyond a 10% increase in program lands in the near future.”
A Review of Real-Time Strategy Game AI (2014, methodology): “An optimal agent is also trying to win at all costs, while the industry really wants game AI that is aiming to lose the game, but in a more human-like way (Davis 1999; Schwab 2013).”
A recipe for short-word pseudo-Anosovs (2010, background): “Mosher has described a unifying approach to both types [Mos], associating to a mapping class f what he calls its active subsurface A(f).”