BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Zachary Fisher

Zachary Fisher contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
hep-thgr-qcArtificial IntelligenceComputation and LanguageMethodology

Trust snapshot

Quick read

Trust 21 - Emerging
9works
0followers
5topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

9 published item(s)

preprint2022arXiv

LogicInference: A New Dataset for Teaching Logical Inference to seq2seq Models

Machine learning models such as Transformers or LSTMs struggle with tasks that are compositional in nature such as those involving reasoning/inference. Although many datasets exist to evaluate compositional generalization, when it comes to evaluating inference abilities, options are more limited. This paper presents LogicInference, a new dataset to evaluate the ability of models to perform logical inference. The dataset focuses on inference using propositional logic and a small subset of first-order logic, represented both in semi-formal logical notation, as well as in natural language. We also report initial results using a collection of machine learning models to establish an initial baseline in this dataset.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Making Transformers Solve Compositional Tasks

Several studies have reported the inability of Transformer models to generalize compositionally, a key type of generalization in many NLP tasks such as semantic parsing. In this paper we explore the design space of Transformer models showing that the inductive biases given to the model by several design decisions significantly impact compositional generalization. Through this exploration, we identified Transformer configurations that generalize compositionally significantly better than previously reported in the literature in a diverse set of compositional tasks, and that achieve state-of-the-art results in a semantic parsing compositional generalization benchmark (COGS), and a string edit operation composition benchmark (PCFG).

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Evaporating Black Holes Coupled to a Thermal Bath

We study the doubly holographic model of [arXiv:1908.10996] in the situation where a black hole in two-dimensional JT gravity theory is coupled to an auxiliary bath system at arbitrary finite temperature. Depending on the initial temperature of the black hole relative to the bath temperature, the black hole can lose mass by emitting Hawking radiation, stay in equilibrium with the bath or gain mass by absorbing thermal radiation from the bath. In all of these scenarios, a unitary Page curve is obtained by applying the usual prescription for holographic entanglement entropy and identifying the quantum extremal surface for the generalized entropy, using both analytical and numeric calculations. As the application of the entanglement wedge reconstruction, we further investigate the reconstruction of the black hole interior from a subsystem containing the Hawking radiation. We examine the roles of the Hawking radiation and also the purification of the thermal bath in this reconstruction.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Information Flow in Black Hole Evaporation

Recently, new holographic models of black hole evaporation have given fresh insights into the information paradox [arXiv:1905.08255, arXiv:1905.08762, arXiv:1908.10996]. In these models, the black hole evaporates into an auxiliary bath space after a quantum quench, wherein the holographic theory and the bath are joined. One particularly exciting development is the appearance of "ER=EPR"-like wormholes in the (doubly) holographic model of [arXiv:1908.10996]. At late times, the entanglement wedge of the bath includes the interior of the black hole. In this paper, we employ both numerical and analytic methods to study how information about the black hole interior is encoded in the Hawking radiation. In particular, we systematically excise intervals from the bath from the system and study the corresponding Page transition. Repeating this process ad infinitum, we end up with a fractal structure on which the black hole interior is encoded, implementing the uberholography protocol of [arXiv:1612.00017].

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

A Quantum Focussing Conjecture

We propose a universal inequality that unifies the Bousso bound with the classical focussing theorem. Given a surface $σ$ that need not lie on a horizon, we define a finite generalized entropy $S_\text{gen}$ as the area of $σ$ in Planck units, plus the von Neumann entropy of its exterior. Given a null congruence $N$ orthogonal to $σ$, the rate of change of $S_\text{gen}$ per unit area defines a quantum expansion. We conjecture that the quantum expansion cannot increase along $N$. This extends the notion of universal focussing to cases where quantum matter may violate the null energy condition. Integrating the conjecture yields a precise version of the Strominger-Thompson Quantum Bousso Bound. Applied to locally parallel light-rays, the conjecture implies a Quantum Null Energy Condition: a lower bound on the stress tensor in terms of the second derivative of the von Neumann entropy. We sketch a proof of this novel relation in quantum field theory.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Proof of the Quantum Null Energy Condition

We prove the Quantum Null Energy Condition (QNEC), a lower bound on the stress tensor in terms of the second variation in a null direction of the entropy of a region. The QNEC arose previously as a consequence of the Quantum Focussing Conjecture, a proposal about quantum gravity. The QNEC itself does not involve gravity, so a proof within quantum field theory is possible. Our proof is somewhat nontrivial, suggesting that there may be alternative formulations of quantum field theory that make the QNEC more manifest. Our proof applies to free and superrenormalizable bosonic field theories, and to any points that lie on stationary null surfaces. An example is Minkowski space, where any point $p$ and null vector $k^a$ define a null plane $N$ (a Rindler horizon). Given any codimension-2 surface $Σ$ that contains $p$ and lies on $N$, one can consider the von Neumann entropy $S_\text{out}$ of the quantum state restricted to one side of $Σ$. A second variation $S_\text{out}^{\prime\prime}$ can be defined by deforming $Σ$ along $N$, in a small neighborhood of $p$ with area $\cal A$. The QNEC states that $\langle T_{kk}(p) \rangle \ge \frac{\hbar}{2π} \lim_{{\cal A}\to 0}S_\text{out}^{ \prime\prime}/{\cal A}$.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

robumeta: An R-package for robust variance estimation in meta-analysis

Meta-regression models are commonly used to synthesize and compare effect sizes. Unfortunately, traditional meta-regression methods are ill-equipped to handle the complex and often unknown correlations among non-independent effect sizes. Robust variance estimation (RVE) is a recently proposed meta-analytic method for dealing with dependent effect sizes. The robumeta package provides functions for performing robust variance meta-regression using both large and small sample RVE estimators under various weighting schemes. These methods are distribution free and provide valid point estimates, standard errors and hypothesis tests even when the degree and structure of dependence between effect sizes is unknown.

0 citations0 reviewsNo code linkedOpen access
preprint2014arXiv

Entropy on a null surface for interacting quantum field theories and the Bousso bound

We study the vacuum-subtracted von Neumann entropy of a segment on a null plane. We argue that for interacting quantum field theories in more than two dimensions, this entropy has a simple expression in terms of the expectation value of the null components of the stress tensor on the null interval. More explicitly $ΔS = 2π\int d^{d-2}y \int_0^1 dx^+\, g(x^+)\, \langle T_{++}\rangle$, where $g(x^+)$ is a theory-dependent function. This function is constrained by general properties of quantum relative entropy. These constraints are enough to extend our recent free field proof of the quantum Bousso bound to the interacting case. This unusual expression for the entropy as the expectation value of an operator implies that the entropy is equal to the modular Hamiltonian, $ΔS = \langle ΔK \rangle $, where $K$ is the operator in the right hand side. We explain how this equality is compatible with a non-zero value for $ΔS$. Finally, we also compute explicitly the function $g(x^+)$ for theories that have a gravity dual.

0 citations0 reviewsNo code linkedOpen access
preprint2014arXiv

Proof of a Quantum Bousso Bound

We prove the generalized Covariant Entropy Bound, $ΔS\leq (A-A')/4G\hbar$, for light-sheets with initial area $A$ and final area $A'$. The entropy $ΔS$ is defined as a difference of von Neumann entropies of an arbitrary state and the vacuum, with both states restricted to the light-sheet under consideration. The proof applies to free fields, in the limit where gravitational backreaction is small. We do not assume the null energy condition. In regions where it is violated, we find that the bound is protected by the defining property of light-sheets: that their null generators are nowhere expanding.

0 citations0 reviewsNo code linkedOpen access