Talks

Quantum gravity has many conceptual problems. Amongst the most well-known is the "Problem of Time": gravitational observables are global in time, while we would really like to obtain probabilities for processes taking us from an observable at one time to another, later one. Tackling these questions using relationalism will be the preferred strategy during this talk. The 'relationalist' approach leads us to shed much redundant information and enables us to identify a reduced configuration space as the arena on which physics unfolds, a goal still beyond our reach in general relativity. Moreover, basing our ontology on this space has far-reaching consequences. One is that it suggests a natural interpretation of quantum mechanics; it is a form of ‘Many-Worlds’ called Many-Instant QM. Another is that the gravitational reduced configuration space has a rich, highly asymmetric structure which singles out preferred, non-singular and homogeneous initial conditions for a wave-function of the universe, which is yet to be explored.   

We establish a direct connection between scattering amplitudes for bi-adjoint scalar theories and a classic polytope--the "associahedron"--known to mathematicians since the 1960s. We find an associahedron naturally living in kinematic space. The tree level scattering amplitude is simply a geometric invariant of the associahedron called its "canonical form" [2], which is a differential form on kinematic space with logarithmic singularities on the boundaries of the associahedron. We show that basic physical principles like locality and unitarity are "rediscovered" as properties of the geometry, and certain "soft" limits can be converted to geometric statements as well.

The associahedron in kinematic space plays an important role in the context of scattering equations. We discover that the scattering equations act as a diffeomorphism between the interior of the open string moduli space (yet another associahedron) and the associahedron in kinematic space. This observation provides the key to a novel derivation of the bi-adjoint CHY formula. Finally, we emphasize the importance of the scattering amplitude as a differential form, as suggested by the "canonical form" construction. We argue on general grounds that every color-dressed amplitude is dual to a form on kinematic space. This is motivated by the surprising observation that "projective" forms on kinematic space satisfy Jacobi identities and therefore have novel implications for color-kinematics duality.

References:

[1] Nima Arkani-Hamed, Yuntao Bai, Song He & Gongwang Yan. In preparation.

[2] Nima Arkani-Hamed, Yuntao Bai & Thomas Lam. Positive Geometries and Canonical Forms. arXiv 1703.04541

The radial acceleration relation is an empirical universal scaling relation between the total gravitational field and the Newtonian acceleration generated by baryons at any given radius within spiral galaxies. In this talk, I will discuss the possibility that such a relation arises from interactions between baryons and dark matter (DM), rather than from feedback processes or modifications of gravity. Starting from this premise, I will discuss what we can infer about the nature of baryon-DM interactions. In particular, I will argue that  such interactions must depend on the local DM density in order to correctly reproduce the observed behavior of spiral and elliptical galaxies. I will briefly revisit existing phenomenological, astrophysical and cosmological constraints on baryon-DM interactions in light of the unusual density dependence of our cross-section. I will conclude by mentioning possible ways of realizing this scenario in particle physics models, and discuss a non-relativistic toy model.

I will describe our recent work from 1709.07460, where we introduce a new renormalization group algorithm for tensor networks. The algorithm is based on a novel understanding of local correlations in a tensor network, and a simple method to remove such correlations from any network. It performs comparably with the best competing algorithms on 2D/(1+1)D systems, but is significantly simpler to implement, and easier to generalize to different lattices and graphs, including to higher dimensions. I will begin the talk by discussing renormalization group methods for tensor networks in general, then describe our algorithm and its advantages, show some benchmark results, and finally comment on the status of implementing real-space renormalization for 3D tensor networks.