Complexity Zoo

Complexity Zoo
Name	Complexity Zoo
Type	Encyclopedia project
Subject	Computational complexity theory
Founded	2005
Founder	Scott Aaronson
Language	English

Complexity Zoo The Complexity Zoo is an online encyclopedia dedicated to cataloging complexity classes in theoretical computer science. It provides systematic entries on decision problems, resource-bounded computation, and structural relationships among classes, serving researchers and students studying topics like Traveling Salesman Problem, P versus NP problem, Quantum computation, Cryptography, and Circuit complexity. The Zoo interrelates classes with results originating from conferences such as STOC, FOCS, ICALP, and institutions including MIT, UC Berkeley, and University of Waterloo.

Overview

The project organizes an expansive taxonomy of complexity classes ranging from classical classes like P (complexity), NP (complexity), co-NP to probabilistic classes such as BPP and interactive classes like IP (complexity), alongside quantum classes like BQP and counting classes like #P. It emphasizes provable inclusions and separations, reductions, complete problems, and oracle results derived from work at Bell Labs, IBM Research, and Microsoft Research. The Zoo links many classes to canonical problems studied at venues such as SODA, CCC (Conference on Computational Complexity), and textbooks by authors affiliated with Princeton University Press and Cambridge University Press.

History and development

Originally compiled in the mid-2000s by a researcher active at MIT and later maintained by contributors from universities such as Harvard University, University of Chicago, and Columbia University, the project grew alongside developments in Quantum information science and advances in structural complexity. Early inspirations included surveys from researchers at Bell Labs and expository lists from workshops at Simons Institute for the Theory of Computing and the Fields Institute. The Zoo expanded as results on interactive proofs at DIMACS workshops, oracle separations published in journals like Journal of the ACM, and breakthroughs concerning derandomization at University of California, Santa Cruz and Caltech accumulated. Over time, entries were augmented to reflect proofs from authors affiliated with Stanford University, Carnegie Mellon University, and contributors who presented at Eurocrypt and Crypto.

Classification and notation

Entries are organized by resource bounds (time, space), computational models (deterministic, nondeterministic, probabilistic, quantum, alternating), and algebraic or counting paradigms. Standard notation follows conventions established in monographs by specialists at Oxford University Press and lecture notes from courses at ETH Zurich, University of Oxford, and University of Cambridge. The Zoo uses many-witness notations linking classes to complete problems such as Boolean satisfiability problem, reductions like polynomial-time many-one reductions studied by researchers at Cornell University, and relativized separations invoking oracles used in work from Princeton University. It tracks closure properties, hierarchies (e.g., the polynomial hierarchy associated with Richard E. Ladner and others), and alternative models introduced in papers from Los Alamos National Laboratory and RIKEN.

Notable complexity classes

The site gives detailed entries on widely studied classes: P (complexity), NP (complexity), co-NP, PSPACE, EXPTIME, NEXP, BPP, RP, ZPP, IP (complexity), AM (complexity), MA (complexity), PH (complexity), #P, Toda's theorem, BQP, QMA, QIP, SZK, UP (complexity), L (complexity), NL (complexity), RL (complexity), AC0, NC (complexity), and classes capturing circuit lower bounds explored at Princeton University. It also covers more specialized or exotic classes introduced in single-author papers from groups at Tel Aviv University, Weizmann Institute of Science, EPFL, Università di Pisa, and Technion. Each class entry cites complete problems, inclusions, oracle separations, and canonical references to original publications in venues like SIAM Journal on Computing and Annals of Mathematics where relevant.

Applications and influence

Beyond cataloging, the Zoo functions as a pedagogical resource used in courses at MIT, Stanford University, UC Berkeley, and University of Toronto. It aids researchers working on hardness results for cryptographic schemes designed by teams at RSA Security, Intel, and academic groups collaborating with National Institute of Standards and Technology. The classification of classes has guided proof techniques in approximation algorithms presented at ICALP and hardness reductions underlying results in parameterized complexity published by researchers at University of Oxford. Quantum class entries have informed experimental and theoretical work in labs at IBM, Google, and University of Maryland focusing on quantum supremacy and algorithmic implications. Policymakers and grant agencies including the National Science Foundation have cited the conceptual clarity offered by systematic taxonomies when prioritizing research programs.

Criticisms and limitations

Critics note that an encyclopedic listing can imply undue equivalence among classes whose separations are unresolved, a concern echoed in debates at panels hosted by Simons Foundation and workshops at Banff International Research Station. The Zoo necessarily lags behind rapidly evolving literature from preprint archives used heavily by groups at Perimeter Institute and Institute for Advanced Study; maintaining up-to-date, peer-reviewed entries is resource-intensive for editors affiliated with University of Chicago. Some commentators from conferences like NeurIPS and ICLR have urged clearer demarcation between empirical complexity in applied domains and theoretical definitional work cataloged by the project. Finally, reliance on conventional notation favored by monographs from Cambridge University Press may obscure alternative formalisms developed at institutions like Rutgers University and University of Edinburgh.

Category:Theoretical computer science