Loglinear Hawkes processes

Jacek Jakubowski; Mariusz Niew\k{e}g{\l}owski; Matthias Kirchner; Tomasz R. Bielecki

REVIEW 1 major objections 4 minor 14 references

Loglinear Hawkes processes

T0 review · 1 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proves that the sign of the memory function near zero decides whether a loglinear Hawkes process explodes, and that nonpositive memory yields stability in distribution with a unique stationary limit.

desk verdict Useful first results for loglinear Hawkes nonexplosion and stability, but the explosion proof has a real gap where the initial condition drops out. read the letter →

arxiv 2507.11265 v1 pith:CFB2EZPG submitted 2025-07-15 math.PR

Tomasz R. Bielecki , Jacek Jakubowski , Matthias Kirchner , Mariusz Niewęg{l}owski This is my paper · ORCID

classification math.PR MSC 60G5560G10

keywords loglinearHawkesprocessnonlinearexplosioncriterionnonexplosionstabilityindistributioninhibitionexponentialratefunctionpoint

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Loglinear Hawkes processes are point processes whose conditional intensity is $\lambda(t)=\exp\{\nu+\int_{(-\infty,t)}h(t-s)N(ds)\}$, the exponential version of a nonlinear Hawkes process. The paper aims to give the first existence, explosion, and stability criteria for this model, which is not covered by earlier Lipschitz-based results because the exponential function is neither Lipschitz nor linearly bounded. Its main theorem shows that the sign of the memory function $h$ near zero decides explosion: if $h\leq 0$ on an initial interval the process is nonexplosive, while if $h>0$ on an initial interval the process is explosive. For nonpositive $h$ with finite first moment, the dynamics are stable in distribution and admit a unique stationary version. A sympathetic reader should care because this is the standard 'spike response' model in computational neuroscience, and the paper supplies the missing theoretical footing for its use and simulation.

What carries the argument

The carrying object is the exponential transfer function $\varphi(x)=e^{\nu+x}$, which maps the integrated past $\int_{(-\infty,t)}h(t-s)N(ds)$ to a positive rate. The construction uses Poisson embedding: take a unit-rate Poisson random measure on $\mathbb{R}_{\ge0}\times\mathbb{R}_{\ge0}$ and keep points whose mark $z$ lies below $\lambda(t)$, so the process is built recursively as $T_n=\inf\{t>T_{n-1}: N(\{t\}\times[0,e^{\nu+\sum_{k<n}h(t-T_k)}])>0\}$. Nonexplosion is proved by bounding $\log\lambda(t)$ on each interval $((k-1)\delta,k\delta]$ by a finite random variable $M_k$, so the number of points per interval is dominated by a Poisson count. Explosion is proved by lower-bounding the conditional probability of a small gap by $1-\exp\{-\frac{\varepsilon}{k^2}e^{\nu+(k-1)h_\delta}\}$, whose product over $k$ is strictly positive exactly when $h_\delta:=\inf_{t\in(0,\delta]}h(t)>0$.

What would settle it

Choose a memory function $h$ that is positive on $(0,\delta]$ but takes large negative values at long lags, and an initial condition $N_0$ with points far in the past, so that $\int_{(-\infty,0]}h(t-r)N_0(dr)$ is finite for every $t>0$ yet tends to $-\infty$ as $t\downarrow0$. Simulate the Poisson-embedding construction of Lemma 3.4 many times and estimate the explosion probability. Theorem 4.1(ii) predicts $P(T_\infty<\infty)>0$ for every such pair satisfying (14); if any such pair fails to explode, the missing uniform lower bound is essential to the theorem as stated.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is Theorem 4.1. For a loglinear Hawkes process with intensity $\lambda(t)=\exp\{\nu+\int_{(-\infty,t)}h(t-s)N(ds)\}$ and initial condition $N_0$ on $(-\infty,0]$, the paper proves: (i) if $\sup_{t>0}h(t)<\infty$, $h(t)\leq0$ for $t\in(0,\delta]$, and $\sup_{t>0}\int_{(-\infty,0]}h(t-r)N_0(dr)<\infty$ a.s., then the process is nonexplosive ($T_\infty=\infty$ a.s.); (ii) if $\inf_{t\in(0,\delta]}h(t)>0$ and $\int_{(-\infty,0]}h(t-r)N_0(dr)>-\infty$ for every $t>0$, then the process is explosive ($T_\infty<\infty$ with positive probability). Theorem 5.2 then shows that for nonpositive $h$ with $\int_0^\infty t|h(t)|dt<\infty$ and a mild decay condition on $N_0$, the process is stable in distribution with a unique stationary version. Theorem 5.3 gives necessary stability conditions, including the universal bound $\int_0^\infty h(s)ds\le e^{-(1+\nu)}$.

Load-bearing premise

The explosion half of the main theorem depends on the unstated assumption that the initial-condition contribution to the log-intensity, $\int_{(-\infty,0]}h(t-r)N_0(dr)$, stays uniformly bounded below for all $t$ in the first window $(0,\delta]$; the theorem's stated condition (14) only says this integral is finite at each individual time, which is weaker and does not by itself justify the uniform lower bound used in the proof.

Editorial extensions

If this is right

If the nonexplosion part is correct, loglinear Hawkes processes with inhibitory memory near zero are legitimate models on the infinite time horizon: almost surely only finitely many events occur in any bounded interval, so simulation and likelihood-based inference are well-defined.
If the explosion part is correct, any memory function that is bounded below by a positive constant on some initial interval produces infinitely many events in finite time with positive probability, so such kernels cannot appear in stationary models.
If the stability theorem is correct, for nonpositive $h$ with $\int_0^\infty t|h(t)|dt<\infty$, the long-run behaviour is insensitive to the initial condition: every solution converges to the same stationary loglinear Hawkes process.
The necessary stability bound $\int_0^\infty h(s)ds\le e^{-(1+\nu)}$ gives a concrete, checkable restriction on any memory kernel that is to yield a stationary version.
The Poisson-embedding construction provides an explicit recursive simulation algorithm for loglinear Hawkes processes, including cases with infinite initial conditions satisfying the theorem's hypotheses.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper's explosion proof in Theorem 4.1(ii) appears to use, without stating it, a uniform lower bound on the initial-condition contribution $\int_{(-\infty,0]}h(t-r)N_0(dr)$ for $t\in(0,\delta]$; condition (14) gives only pointwise finiteness. A natural test is whether processes with memory functions that have negative spikes at large lags and initial points far back still explode, as the theor
The stability result suggests a symmetry with linear Hawkes theory: just as linear Hawkes processes require nonnegative memory for their rate to stay positive, loglinear Hawkes processes require nonpositive memory for stability because the exponential turns any positive excursion into explosive feedback.
The universal bound $\int_0^\infty h(s)ds\le e^{-(1+\nu)}$ could be used as a model-checking diagnostic: estimated memory functions from stable neural spike trains should satisfy this integral constraint, and violations would indicate that the fitted loglinear model cannot be stationary.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Take a look at this one. The genuinely new content: for Hawkes processes with rate exp(nu + integral h dN), they give the first nonexplosion and explosion criteria and a stability result for nonpositive h. Nonexplosion theorem (4.1(i)) is well argued: local negativity of h on (0,delta] plus a one-sided bound on the initial-condition integral controls each interval. The reduction of stability for h<=0 to Bremaud-Massoulie via a bounded Lipschitz truncation is sensible, and the necessary condition via Jensen is clean.

The soft spot is Theorem 4.1(ii), the explosion criterion. Equation (20) conditions on G_{k-1}, which includes N0 and T_1...T_{k-1}, but the exponential in the conditional probability sums only over j=1,...,k-1. The initial-condition integral A(t)=int_{(-inf,0]} h(t-s)N0(ds) is missing. The hypotheses only give A(t)>-inf pointwise in t, not uniformly over t in (0,delta]. The stress-test example is valid: negative spikes at lags around 1/n make A(t) -> -inf as t decreases to 0 while (14) still holds. With A(t) arbitrarily negative near zero, the lower bound exp(nu + (k-1)h_delta) used in (20)-(21) does not follow, and the infinite product can vanish. The theorem may still be true; I would not be surprised if a different argument works, but the proof as written does not establish it. Fixing it likely requires a stronger condition (finite N0, or a uniform lower bound on A(t)) or a substantially different proof.

The rest is in decent shape. The construction in Lemma 3.4 includes the initial condition through the sum over all previous points, so the issue is confined to the proof of (ii). I could not independently verify that the hypotheses of Bremaud-Massoulie Theorem 2 are exactly met, but nothing in the reduction looks off.

Who is this for: people working on nonlinear Hawkes processes, computational neuroscience models, and point-process stability. The nonexplosion and stability-for-inhibition parts are citable once the explosion proof is sorted. I would not cite it in its present form. I would send it to a serious referee; the topic is important and most of the paper is solid, but the referee should focus on Theorem 4.1(ii). A major revision is warranted.

Referee Report

1 major / 4 minor

Summary. The paper introduces loglinear Hawkes processes, a nonlinear Hawkes model with conditional intensity lambda(t)=exp(nu + integral h(t-s) dN(s)), motivated by spike-response models in computational neuroscience. The authors construct the process via a Poisson-embedding/thinning argument, prove a uniqueness statement for the construction, and then establish sufficient conditions for nonexplosion and explosion in Theorem 4.1. For nonexplosion they assume a memory function whose values on (0,delta] are nonpositive and a bounded initial-condition contribution; for explosion they assume positive values on (0,delta] and pointwise finiteness of the initial-condition contribution. In Section 5 they prove stability in distribution for nonpositive memory functions with finite first absolute moment, using a reduction to Brémaud and Massoulié (1996), and derive necessary conditions for stability in Theorem 5.3. The paper closes with an outlook on extensions to partly positive memory functions and networks.

Significance. If the main theorems are correct, the paper fills a real gap: the exponential rate function is neither Lipschitz nor linearly dominated, so existing nonlinear-Hawkes stability and existence results do not apply. The construction is explicit and machine-checkable in style, with a clean Poisson-embedding proof and a clear uniqueness argument. The nonexplosion theorem is carefully argued, and the stability theorem for nonpositive h is a clean, direct application of a classical result with explicit hypotheses. The necessary condition (24) is a simple, falsifiable bound. The paper would be a useful theoretical foundation for loglinear Hawkes models in neuroscience and related fields. However, the proof of the explosion theorem contains a load-bearing gap, described below, that must be repaired before the result can be accepted as stated.

major comments (1)

[Theorem 4.1(ii), equations (17)–(21)] The proof of explosion drops the initial-condition term in the conditional probability calculation. In equation (20), the exponent contains only sum_{j=1}^{k-1} h(t-T_j), but the intensity in (11) is exp(nu + integral_{(-infinity,0]} h(t-s) N0(ds) + sum_{j=1}^{k-1} h(t-T_j)). The lower bound from (20) to (21) is therefore valid only if A(t)=integral_{(-infinity,0]} h(t-s) N0(ds) is uniformly bounded below on the interval (T_{k-1}, T_{k-1}+epsilon/k^2], which is contained in (0,delta] on the event B_{k-1}. Condition (14) only asserts pointwise finiteness for each fixed t>0; it does not imply a uniform lower bound. For instance, with N0=sum_n delta_{-A_n} and h(t)=1 on (0,delta] but h(t)=-n on [A_n+1/(n+1), A_n+1/n] for A_n tending to infinity, (14) holds for each fixed t>0, yet A(t) tends to -infinity as t approaches 0 from above. The infinite product in (21) can then be zero, so the claimed lower bound is unjustified. A repair requires either a uniform lower bound on A(t) over (0,delta], a finite initial condition, or an entirely different argument.

minor comments (4)

[Section 2] There are small typos: 'choise' should be 'choice' and 'n (7)' should be 'in (7)'.
[Lemma 3.4 and surrounding text] The same symbol N is used for the constructed point process, the initial-condition counting measure N0, and the auxiliary Poisson random measure in the embedding; a brief notation remark would improve readability.
[Section 4, Example (iv)] The display involving the dominating function tilde h is garbled in the manuscript and should be reformatted.
[Theorem 4.1(ii), condition (14)] The condition is written as an inequality with '-infinity' without specifying 'almost surely' or 'for every t>0'; adding 'a.s.' and making the quantifier explicit would remove ambiguity, especially because the proof requires a uniform-in-t version.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper derives existence, explosion, and stability results from explicit hypotheses and external theorems, with no fitted parameters or self-referential reductions.

full rationale

The paper's derivation chain is self-contained in the relevant sense. The explosion and nonexplosion criteria in Theorem 4.1 are proved directly from the Poisson embedding construction (Lemma 3.4) and explicit bounds on the conditional intensity; the assumptions (13) and (14) concern the initial condition and memory function and are used as hypotheses, not as disguised conclusions. The stability result in Theorem 5.2 is explicitly presented as an application of Brémaud and Massoulié (1996, Theorem 2), an external theorem, and the proof verifies its hypotheses (nonpositive h, integrability of t|h(t)|, and the decay condition on N0). Theorem 5.3 derives necessary stability conditions from Jensen's inequality and stationarity, again with no parameter fitted to the target conclusion. There are no fitted inputs renamed as predictions, no uniqueness theorem imported from the authors' own prior work, and no ansatz smuggled in via self-citation. The reviewer-identified gap in the proof of Theorem 4.1(ii), concerning the uniform lower bound of the initial-condition integral in equation (20), is a potential correctness flaw, not a circularity: the proof attempts to use the hypotheses and fails to justify a step, but it does not assume the conclusion or define a quantity in terms of the quantity it claims to derive. Since the instructions reserve circularity findings for reductions by construction or fitted parameters renamed as predictions, the appropriate verdict is no significant circularity with score 0.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper's results are analytic; no parameters are fitted and no new entities are postulated. The main external inputs are the Brémaud and Massoulié stability theorems and standard point-process tools. The ad hoc assumption in the explosion proof is the weakest point: it is not implied by the theorem's hypotheses and is the source of the main soundness concern.

assumptions (5)

standard math Poisson embedding construction (Lemma 3.2): a point process with predictable intensity can be represented as a thinning of a homogeneous Poisson random measure.
Used to construct the loglinear Hawkes process in Lemma 3.4; quoted from Brémaud and Massoulié (1996, Lemma 3).
standard math Brémaud and Massoulié (1996, Theorem 2) gives stability in variation for nonlinear Hawkes processes with bounded Lipschitz rate function and integrable memory kernel.
Basis for Theorem 5.2; the paper states all hypotheses are fulfilled but does not restate the theorem's exact conditions, in particular whether the Lipschitz constant must be small relative to the L1 norm of the memory kernel.
domain assumption For a nonpositive memory function h, the argument ∫ h dN is nonpositive, so the truncated rate function φ(x)=e^{ν+x} for x≤0 and e^{ν} for x>0 matches the loglinear intensity.
Used in Theorem 5.2 to apply Brémaud and Massoulié Theorem 2; this requires h(t) ≤ 0 for all t>0.
ad hoc to paper In the explosion proof of Theorem 4.1(ii), the initial-condition term ∫_{(-∞,0]} h(t-s)N0(ds) is either absent from the intensity or uniformly bounded below on the interval (0,δ].
Equation (20) uses only ∑_{j<k} h(t-T_j) for the conditional gap probability; condition (14) only gives pointwise finiteness, not a uniform lower bound, so this assumption is not justified by the stated hypotheses.
standard math Stationarity and Campbell's formula: for a stationary point process with finite mean intensity e^c, E∫ f(s) N(ds) = e^c ∫ f(s) ds.
Used in Theorem 5.3 to derive the necessary stability inequalities (23) and (24).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Loglinear Hawkes processes." pith.science (2026). https://pith.science/paper/CFB2EZPG

@misc{pith2026250711265,
  author       = {Pith},
  title        = {Pith review of: Loglinear Hawkes processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CFB2EZPG}},
  note         = {Machine review of arXiv:2507.11265}
}

read the original abstract

This paper discusses a special class of nonlinear Hawkes processes, where the rate function is the exponential function. We call these processes loglinear Hawkes processes. In the main theorem, we give sufficient conditions for explosion and nonexplosion that cover a large class of practically relevant memory functions. We also investigate stability. In particular, we show that for nonpositive memory functions the loglinear Hawkes process is stable. The paper aims at providing a theoretical basis for further research and applications of these processes.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 13 canonical work pages

[6]

Fluctuations and moderate deviations for the mean fields of Hawkes processes

Fluctuations and moderate de viations for the mean ﬁelds of Hawkes processes. arXiv:2307.15903. Gerhard, F., Deger, M., Truccolo, W.,

work page Pith review arXiv
[13]

Stochastic Processes and their Applicat ions 149, 39–59

Mean-ﬁeld limit of age and leaky memory de pendent hawkes processes. Stochastic Processes and their Applicat ions 149, 39–59. doi:10.1016/j.spa.2022.03.014. Truccolo, W.,

work page doi:10.1016/j.spa.2022.03.014 2022
[31]

URL: http://infoscience.epfl.ch/record/226943, doi:https://doi.org/10.1371/journal.pcbi.1005390

e1005390. URL: http://infoscience.epfl.ch/record/226943, doi:https://doi.org/10.1371/journal.pcbi.1005390. Gerstner, W., van Hemmen, J.L.,

work page doi:10.1371/journal.pcbi.1005390
[1951]

Theory and application of inﬁnite series. T ransl. from the 2nd ed. and revised in accordance with the fourth by R. C. H. Young. Londo n-Glasgow: Blackie & Son, Ltd. XII, 563 p. (1951). Massoulié, L.,

work page 1951
[1992]

Network: Computation in Neural Syste ms 3, 139–164

Associative memory in a network of ’spiking’ neurons. Network: Computation in Neural Syste ms 3, 139–164. doi:10.1088/0954-898X/3/2/002. 16 Gerstner, W., Kistler, W.M., Naud, R., Paninski, L.,

work page doi:10.1088/0954-898x/3/2/002
[1998]

Stochastic Processes and thei r Applications 75, 1–30

Stability results for a general class o f interacting point processes dynamics, and applications. Stochastic Processes and thei r Applications 75, 1–30. URL: https://www.sciencedirect.com/science/article/pii/S0304414998000064, doi:https://doi.org/10.1016/S0304-4149(98)00006-4 . Mei, H., Eisner, J.,

work page doi:10.1016/s0304-4149(98)00006-4
[2006]

Journal of Computational Neuroscience 21, 35–49

P redicting spike timing of neocortical pyramidal neurons by simple threshold models. Journal of Computational Neuroscience 21, 35–49. doi: 10.1007/s10827-006-7074-5 . Knopp, K.,

work page doi:10.1007/s10827-006-7074-5
[2012]

PLOS Computa tional Biology 8, e1002711

Coding and decoding with adapt ing neurons: A population approach to the peri-stimulus time histogram. PLOS Computa tional Biology 8, e1002711. doi:10.1371/journal.pcbi.1002711. Pillow, J.W., Shlens, J., Paninski, L., Sher, A., Litke, A.M ., Chichilnisky, E.J., Simoncelli, E.P.,

work page doi:10.1371/journal.pcbi.1002711

Show all 14 references

[2016]

Journal of physiology, Paris 110, 336–347

From point process observations to coll ective neural dynamics: Non- linear Hawkes process GLMs, low-dimensional dynamics and c oarse graining. Journal of physiology, Paris 110, 336–347. doi: 10.1016/j.jphysparis.2017.02.004. 17 Truccolo, W., Eden, U.T., Fellows, M.R., Donoghu...

2017 doi
[2017]

Stochastic Processes and their Applicatio ns 127, 1840–1869

Multi-class oscillat ing systems of inter- acting neurons. Stochastic Processes and their Applicatio ns 127, 1840–1869. URL: http://www.sciencedirect.com/science/article/pii/S0304414916301739, doi:https://doi.org/10.1016/j.spa.2016.09.013. Duarte, A., Löcherbach, E., Ost, G.,

doi
[2019]

ESAIM: PS 23, 770–796

Stability, conve rgence to equilibrium and simulation of non-linear Hawkes processes with memory kern els given by the sum of erlang kernels. ESAIM: PS 23, 770–796. URL: https://doi.org/10.1051/ps/2019005, doi:10.1051/ps/2019005. Fonte, C., Schmutz, V.,
[2020]

Advances in Applied Proba bility 52, 879–915

Renew al in Hawkes processes with self-excitation and inhibition. Advances in Applied Proba bility 52, 879–915. Daley, D., Vere-Jones, D., 2003/2009. An Introduction to th e Theory of Point Processes. volume I and II. 2nd ed., Springer, New York. Deutsch, I., Ross, G.J.,

2003
[2022]

SIAM Journal on Mathematica l Analysis 54, 4721–4756

Long time behavior of an age- an d leaky memory-structured neuronal population equation. SIAM Journal on Mathematica l Analysis 54, 4721–4756. doi:10.1137/21M1428571. Gao, F., Gao, Y., Zhu, L.,

doi
[2023]

arXiv:2201.05009

Estimating product cannibal isation in wholesale using mul- tivariate Hawkes processes with inhibition. arXiv:2201.05009. Ditlevsen, S., Löcherbach, E.,

arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [6]

Fluctuations and moderate deviations for the mean fields of Hawkes processes

Fluctuations and moderate de viations for the mean ﬁelds of Hawkes processes. arXiv:2307.15903. Gerhard, F., Deger, M., Truccolo, W.,

work page Pith review arXiv

[2] [13]

Stochastic Processes and their Applicat ions 149, 39–59

Mean-ﬁeld limit of age and leaky memory de pendent hawkes processes. Stochastic Processes and their Applicat ions 149, 39–59. doi:10.1016/j.spa.2022.03.014. Truccolo, W.,

work page doi:10.1016/j.spa.2022.03.014 2022

[3] [31]

URL: http://infoscience.epfl.ch/record/226943, doi:https://doi.org/10.1371/journal.pcbi.1005390

e1005390. URL: http://infoscience.epfl.ch/record/226943, doi:https://doi.org/10.1371/journal.pcbi.1005390. Gerstner, W., van Hemmen, J.L.,

work page doi:10.1371/journal.pcbi.1005390

[4] [1951]

Theory and application of inﬁnite series. T ransl. from the 2nd ed. and revised in accordance with the fourth by R. C. H. Young. Londo n-Glasgow: Blackie & Son, Ltd. XII, 563 p. (1951). Massoulié, L.,

work page 1951

[5] [1992]

Network: Computation in Neural Syste ms 3, 139–164

Associative memory in a network of ’spiking’ neurons. Network: Computation in Neural Syste ms 3, 139–164. doi:10.1088/0954-898X/3/2/002. 16 Gerstner, W., Kistler, W.M., Naud, R., Paninski, L.,

work page doi:10.1088/0954-898x/3/2/002

[6] [1998]

Stochastic Processes and thei r Applications 75, 1–30

Stability results for a general class o f interacting point processes dynamics, and applications. Stochastic Processes and thei r Applications 75, 1–30. URL: https://www.sciencedirect.com/science/article/pii/S0304414998000064, doi:https://doi.org/10.1016/S0304-4149(98)00006-4 . Mei, H., Eisner, J.,

work page doi:10.1016/s0304-4149(98)00006-4

[7] [2006]

Journal of Computational Neuroscience 21, 35–49

P redicting spike timing of neocortical pyramidal neurons by simple threshold models. Journal of Computational Neuroscience 21, 35–49. doi: 10.1007/s10827-006-7074-5 . Knopp, K.,

work page doi:10.1007/s10827-006-7074-5

[8] [2012]

PLOS Computa tional Biology 8, e1002711

Coding and decoding with adapt ing neurons: A population approach to the peri-stimulus time histogram. PLOS Computa tional Biology 8, e1002711. doi:10.1371/journal.pcbi.1002711. Pillow, J.W., Shlens, J., Paninski, L., Sher, A., Litke, A.M ., Chichilnisky, E.J., Simoncelli, E.P.,

work page doi:10.1371/journal.pcbi.1002711