Universalité pour des polymères aléatoires. Paris, 17 Novembre 2015

Σχετικά έγγραφα
Comportement critique du modèle d accrochage de. Toulouse, 29 Mars 2016

HOW DOES A CHARGED POLYMER COLLAPSE?

P AND P. P : actual probability. P : risk neutral probability. Realtionship: mutual absolute continuity P P. For example:

The Probabilistic Method - Probabilistic Techniques. Lecture 7: The Janson Inequality

Survival Analysis: One-Sample Problem /Two-Sample Problem/Regression. Lu Tian and Richard Olshen Stanford University

Statistical Inference I Locally most powerful tests

Overview. Transition Semantics. Configurations and the transition relation. Executions and computation

Fractional Colorings and Zykov Products of graphs

4.6 Autoregressive Moving Average Model ARMA(1,1)

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

Partial Differential Equations in Biology The boundary element method. March 26, 2013

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Uniform Convergence of Fourier Series Michael Taylor

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

From the finite to the transfinite: Λµ-terms and streams

Lecture 2. Soundness and completeness of propositional logic

Nowhere-zero flows Let be a digraph, Abelian group. A Γ-circulation in is a mapping : such that, where, and : tail in X, head in

Approximation of distance between locations on earth given by latitude and longitude

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8 questions or comments to Dan Fetter 1

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Elements of Information Theory

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

Homework 8 Model Solution Section

Abstract Storage Devices

Homework 3 Solutions

Lecture 34 Bootstrap confidence intervals

Example Sheet 3 Solutions

Instruction Execution Times

g-selberg integrals MV Conjecture An A 2 Selberg integral Summary Long Live the King Ole Warnaar Department of Mathematics Long Live the King

Chap. 6 Pushdown Automata

Theorem 8 Let φ be the most powerful size α test of H

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

ST5224: Advanced Statistical Theory II

A Bonus-Malus System as a Markov Set-Chain. Małgorzata Niemiec Warsaw School of Economics Institute of Econometrics

Other Test Constructions: Likelihood Ratio & Bayes Tests

Local Approximation with Kernels

EE 570: Location and Navigation

Mean-Variance Analysis

EE512: Error Control Coding

Arithmetical applications of lagrangian interpolation. Tanguy Rivoal. Institut Fourier CNRS and Université de Grenoble 1

Markov chains model reduction

Numerical Analysis FMN011

Major Concepts. Multiphase Equilibrium Stability Applications to Phase Equilibrium. Two-Phase Coexistence

Second Order RLC Filters

Heisenberg Uniqueness pairs

ANNEALED SCALING FOR A CHARGED POLYMER

Estimation for ARMA Processes with Stable Noise. Matt Calder & Richard A. Davis Colorado State University

The Simply Typed Lambda Calculus

Iterated trilinear fourier integrals with arbitrary symbols

A Lambda Model Characterizing Computational Behaviours of Terms

Reflecting Brownian motion in two dimensions: Exact asymptotics for the stationary distribution

2 Composition. Invertible Mappings

The circle theorem and related theorems for Gauss-type quadrature rules

12. Radon-Nikodym Theorem

ω ω ω ω ω ω+2 ω ω+2 + ω ω ω ω+2 + ω ω+1 ω ω+2 2 ω ω ω ω ω ω ω ω+1 ω ω2 ω ω2 + ω ω ω2 + ω ω ω ω2 + ω ω+1 ω ω2 + ω ω+1 + ω ω ω ω2 + ω

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Math 446 Homework 3 Solutions. (1). (i): Reverse triangle inequality for metrics: Let (X, d) be a metric space and let x, y, z X.

5. Choice under Uncertainty

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

Optimal Impartial Selection

Probability and Random Processes (Part II)

Aquinas College. Edexcel Mathematical formulae and statistics tables DO NOT WRITE ON THIS BOOKLET

Affine Weyl Groups. Gabriele Nebe. Summerschool GRK 1632, September Lehrstuhl D für Mathematik

Monte Carlo Methods. for Econometric Inference I. Institute on Computational Economics. July 19, John Geweke, University of Iowa

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

MINIMAL CLOSED SETS AND MAXIMAL CLOSED SETS

Depth versus Rigidity in the Design of International Trade Agreements. Leslie Johns

Exercises 10. Find a fundamental matrix of the given system of equations. Also find the fundamental matrix Φ(t) satisfying Φ(0) = I. 1.

Inverse trigonometric functions & General Solution of Trigonometric Equations

ΕΙΣΑΓΩΓΗ ΣΤΗ ΣΤΑΤΙΣΤΙΚΗ ΑΝΑΛΥΣΗ

Bayesian modeling of inseparable space-time variation in disease risk

Math221: HW# 1 solutions

Homework for 1/27 Due 2/5

C.S. 430 Assignment 6, Sample Solutions

Reminders: linear functions

LAD Estimation for Time Series Models With Finite and Infinite Variance

New bounds for spherical two-distance sets and equiangular lines

Every set of first-order formulas is equivalent to an independent set

The challenges of non-stable predicates

ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =?

Lecture 21: Properties and robustness of LSE

ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ. Ψηφιακή Οικονομία. Διάλεξη 7η: Consumer Behavior Mαρίνα Μπιτσάκη Τμήμα Επιστήμης Υπολογιστών

Sequent Calculi for the Modal µ-calculus over S5. Luca Alberucci, University of Berne. Logic Colloquium Berne, July 4th 2008

Problem Set 3: Solutions

6.3 Forecasting ARMA processes

MATH423 String Theory Solutions 4. = 0 τ = f(s). (1) dτ ds = dxµ dτ f (s) (2) dτ 2 [f (s)] 2 + dxµ. dτ f (s) (3)

ΠΑΝΕΠΙΣΤΗΜΙΟ ΠΕΙΡΑΙΑ ΤΜΗΜΑ ΝΑΥΤΙΛΙΑΚΩΝ ΣΠΟΥΔΩΝ ΠΡΟΓΡΑΜΜΑ ΜΕΤΑΠΤΥΧΙΑΚΩΝ ΣΠΟΥΔΩΝ ΣΤΗΝ ΝΑΥΤΙΛΙΑ

Credit Risk. Finance and Insurance - Stochastic Analysis and Practical Methods Spring School Jena, March 2009

TMA4115 Matematikk 3

( y) Partial Differential Equations

ΠΤΥΧΙΑΚΗ ΕΡΓΑΣΙΑ "ΠΟΛΥΚΡΙΤΗΡΙΑ ΣΥΣΤΗΜΑΤΑ ΛΗΨΗΣ ΑΠΟΦΑΣΕΩΝ. Η ΠΕΡΙΠΤΩΣΗ ΤΗΣ ΕΠΙΛΟΓΗΣ ΑΣΦΑΛΙΣΤΗΡΙΟΥ ΣΥΜΒΟΛΑΙΟΥ ΥΓΕΙΑΣ "

Graded Refractive-Index

1. (a) (5 points) Find the unit tangent and unit normal vectors T and N to the curve. r(t) = 3cost, 4t, 3sint

Phase-Field Force Convergence

ECE Spring Prof. David R. Jackson ECE Dept. Notes 2

FORMULAS FOR STATISTICS 1

Models for Probabilistic Programs with an Adversary

Transcript:

Universalité pour des polymères aléatoires Niccolò Torri Paris, 17 Novembre 2015

1 Polymers 2 Pinning Model and DPRE 3 Universality of the Pinning Model New Results 4 Proofs Continuum model Proof of our results N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 1 / 26

Polymers Interactions with Itself External environment Depend on some parameters Aim Spatial configuration Phase transition? Critical Points? N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 2 / 26

Intermezzo: Some Basic Probabilistic Processes Random Walk on Z d S n = n i=1 X i, X i = increments. x i { Self Avoiding Random Walk (SAW) on Z d Conditioned to visit at most ones each state. N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 3 / 26

Abstract Monomers and Abstract Polymers Increment X i a monomer Use SAW: Abstract Polymer (N monomers): N-increments of a SAW N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 4 / 26

Abstract Monomers and Abstract Polymers Increment X i a monomer Use SAW: SAW: challenging object! Subclass: Directed Random Walks (Directed Polymers) Abstract Polymer (N monomers): N-increments of a SAW s n n N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 4 / 26

1 Polymers 2 Pinning Model and DPRE 3 Universality of the Pinning Model New Results 4 Proofs Continuum model Proof of our results

Interaction with the environment Pinning Model interaction with a membrane Directed Polymer in Random Environment (DPRE) interaction with the whole environment s n n region of interaction : straight line N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 5 / 26

Pinning Model s n n region of interaction : straight line Region of interaction Lattice Ω = {1,, N} Contact process Renewal Process τ = {n : S n = 0} N 0 Interaction P(τ i τ i 1 = n) P ω N, β, h (τ) = 1 Z ω N,β,h c n 1+α, α > 0 exp (βω i + h)1 {i τ} P(τ) i Ω N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 6 / 26

Directed Polymer in Random Environment (DPRE) Region of interaction Lattice Ω = {1,, N} Z sn =X1+... + Xn Polymer Directed Walk (n, S n ) n N Ω Interaction P ω N, β, h (S) = 1 Z ω exp (βω (i,x) + h)1 {Si =x} N,β,h P(S) (i,x) Ω x p(x,x+1) = 1 2 p(x,x - 1) = 1 2 n N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 7 / 26

1 Polymers 2 Pinning Model and DPRE 3 Universality of the Pinning Model New Results 4 Proofs Continuum model Proof of our results

Behavior of the model Pinning Model: interaction polymer and membrane. Renewal Process τ N 0 (τ i τ i 1 ) i N i.i.d. & P(τ i τ i 1 = n) c n 1+α, α > 0 Pinning Model τ [0, N] perturbed P ω N, β, h (τ) = 1 N Z ω exp (βω i + h)1 {i τ} P(τ) N,β,h disorder (quenched realisation of) (ω = (ω x ) x Ω, P) i.i.d. E(ω 1 ) = 0, Var(ω 1 ) = 1, Λ(z) = log E(e zω 1 ) <, z small. i=1 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 8 / 26

Behavior of the model Pinning Model: interaction polymer and membrane. Renewal Process τ N 0 (τ i τ i 1 ) i N i.i.d. & P(τ i τ i 1 = n) c n 1+α, α > 0 Pinning Model τ [0, N] perturbed P ω N, β, h (τ) = 1 N Z ω exp (βω i + h)1 {i τ} P(τ) N,β,h disorder (quenched realisation of) (ω = (ω x ) x Ω, P) i.i.d. E(ω 1 ) = 0, Var(ω 1 ) = 1, Λ(z) = log E(e zω 1 ) <, z small. i=1 Goal: Study τ when N gets large. τ [0, N] c h,β N or τ [0, N] = o(n)? Phase Transition? Critical point? N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 8 / 26

Localization/Delocalization Free Energy F (α) 1 (β, h) = lim N N E [ ] log Z ω β,h,n ( ) {points of τ N} h F (α) (β, h) = lim EE ω N β,h,n N ( ) β 0 there exists a critical point h c (β) s n Delocalized Localized n h > h c (β) localization, ( ) > 0, h < h c (β) de-localization, ( ) = 0. N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 9 / 26

Analysis of the model Goal: understand h c (β). Homogeneous model h c (0) explicit. It provides Lower/Upper bounds h c (0) Λ(β) h c (β) < h c (0), h c (0) Λ(β) annealed critical point. N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 10 / 26

Analysis of the model Goal: understand h c (β). Homogeneous model h c (0) explicit. It provides Lower/Upper bounds h c (0) Λ(β) h c (β) < h c (0), h c (0) Λ(β) annealed critical point. P ω N, β, h (τ) = 1 Z ω N,β,h N exp (βω i + h)1 {i τ} P(τ) i=1 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 10 / 26

Relevance/Irrelevance of the disorder α > 1/2 relevant disorder 0 < α < 1/2 irrelevant disorder h c(β) > h c(0)-λ(β) β>0 h c(β) = h c(0)-λ(β) if β small h (β) c h (0) c h (0)-Λ(β) c N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 11 / 26

Relevance/Irrelevance of the disorder α > 1/2 relevant disorder h c(β) > h c(0)-λ(β) β>0 h (β) c h (0) c h (0)-Λ(β) c Aim: When α > 1/2, asymptotics of h c (β) as β 0 (Weak Disorder). N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 11 / 26

An overview of the literature case α > 1 Theorem (Q. Berger, F. Caravenna, J. Poisat, R. Sun, N. Zygouras, 2014) Let α > 1, then h c (β) β 0 cβ 2, where c is explicit depending on the law of τ i τ i 1. case α (1/2, 1) several authors K. S. Alexander, B. Derrida, G. Giacomin, H. Lacoin, F. L. Toninelli and N. Zygouras (2008 2011): Theorem Let α (1/2, 1), then there exist 0 < c < C < such that for β small. c β 2α 2α 1 hc (β) C β 2α 2α 1, N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 12 / 26

3 Universality of the Pinning Model New Results

Results Theorem (Caravenna, Toninelli, T., 2015) Universal feature of the Free Energy: α (1/2, 1) F (α) F ( ˆβ, (α)( ˆβ ε α 1 2, ĥ ε α) ĥ) = lim ε 0 ε Rescale relation F (α) ( ˆβ, ĥ) = β ( ) 2 2α 1 F (α) 1, ĥβ 2α 2α 1 ĉ universal constant depending on α and c. H (α) ( ˆβ) = ĉ ˆβ 2α 2α 1 Theorem (Caravenna, Toninelli, T., 2015) Universal Critical Behavior h c (β) β 0 ĉ β 2α 2α 1, N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 13 / 26

Conjectures for DPRE Universal feature of the Free Energy F DPRE F (α)( ε (1) = lim ε 0 ε 1 ) 4 F (α)( ε ) ε 4 F DPRE (1). Conjecture supported by Lacoin (2009) (no sharp result) N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 14 / 26

4 Proofs Continuum model Proof of our results

General Framework Discrete Lattice Ω = Ω Z d, Reference System (σ = (σ x ) x Ω, P Ω ), σ x {0, 1}, Disorder (quenched realisation of) (ω = (ω x ) x Ω, P) i.i.d. E(ω 1 ) = 0, Var(ω 1 ) = 1, Λ(z) = log E(e zω 1 ) <, z small. Interaction (disordered measure) P ω Ω,β,h (dσ) = 1 Z ω exp (βω i + h)σ x P Ω(dσ) N,β,h x Ω N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 15 / 26

General Framework Discrete Lattice Ω = Ω Z d, Reference System (σ = (σ x ) x Ω, P Ω ), σ x {0, 1}, Interaction (disordered measure) P ω Ω,β,h (dσ) = 1 Z ω exp (βω i + h)σ x P Ω(dσ) N,β,h x Ω Pinning Model DPRE Lattice Ω = (0, N) Ω = (0, N) R Reference System σ x = 1 {x τ} σ (n,x) = 1 {Sn =x} N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 15 / 26

General Framework: Goal Pinning Model DPRE Lattice Ω = (0, N) Ω = (0, N) R Reference System σ x = 1 {x τ} σ (n,x) = 1 {Sn =x} Assumption: Continuum Limit rescaling δ = δ N of Ω s.t. (σ = (σ x ) x Ωδ, P Ωδ ) δ 0 ( ˆσ = ( ˆσ x ) x Ω, ˆP Ω ) Pinning Model DPRE Rescaling δ = 1 N δ = ( 1 N, 1 ) N 1/2 Continuum limit Regenerative set ˆτ Brownian Motion (B t ) t Continuum system ˆσ x = 1 {x ˆτ} ˆσ x = 1 {Bt =y}, N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 16 / 26

General Framework: Goal Convergence of reference model: (σ = (σ x ) x Ωδ, P Ωδ ) δ 0 ( ˆσ = ( ˆσ x ) x Ω, ˆP Ω ) Convergence of the disordered model: rescaling of β, h = β, h(δ) δ 0 0: (σ = (σ x ) x Ωδ, P ω Ω δ,β δ,h δ ) δ 0? Idea look at the partition function Z ω Ω,β,h...Because it defines completely P ω Ω,β,h N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 16 / 26

Chaos Expansion Partition function Z ω Ω,β,h = E [ ] Ω e n Ω(βω n +h)σ n Polynomial Chaos Expansion where Z ω Ω,β,h = E Ω (1 + ε n σ n ) = n Ω Ω 1 1 + ϕ (k) k! Ω (n 1,, n k ) k=1 (n 1,,n k ) Ω k ε n = e (βω n+h) 1 ϕ (k) Ω (n 1,, n k ) = E Ω [ σn1 σ nk ] k k=1 ε ni N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 17 / 26

Continuum Model Theorem (Caravenna, Sun, Zygouras (2014)) Assumption γ : (δ γ ) k ϕ (k) Ω δ (x 1,, x k ) δ 0 ˆϕ (k) Ω (x 1,, x k ). Choosing β δ = ˆβδ d/2 γ, h δ = ĥδ d γ Z ω Ω δ,β δ,h δ (d) δ 0 ẐW Ω, ˆβ,ĥ Continuum Partition function: Wiener Chaos Expansion Ẑ W Ω, ˆβ,ĥ = 1 + 1 k! k=1 (k) ˆϕ (x 1,,x k ) Ω k Ω (x 1,, x k ) k [ ] ˆβW(dx i ) + ĥdx i k=1 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 18 / 26

Proof Ω δ Z ω Ω δ,β,h = 1 + k=1 1 k! (x 1,,x k ) Ω k δ Idea Lindeberg Principle: ε x is quite Gaussian : ϕ (k) Ω δ (x 1,, x k ) k k=1 ε xi ε x = e βω x+h 1 : Var(ε x ) β 2, E(ε x ) h + β2 2 = h : ε x βw 1 + h = β x d/2 W x + h (x 1,,x k ) Ω k δ ϕ (k) Ω δ (x 1,, x k ) (k) ϕ (x 1,,x k ) Ω k k k=1 ε xi Ω δ (x 1,, x k ) k [ ] βδ d/2 W(dx i ) + h δ d dx i k=1 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 19 / 26

Example: Pinning Model Polynomial Chaos expansion: ϕ (k) Ω (n [ ] [ ] 1,, n k ) = E Ω σn1 σ nk = EΩ 1n1 τ 1 nk τ Theorem (Doney (1997)) P(τ 1 = n) c, α (0, 1) P(n τ) c n 1+α n 1 α N Z ω Ω,β,h 1 + k=1 0<n 1 < <n k <N C k n 1 α 1 (n k n k 1 ) 1 α k k=1 ε xi N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 20 / 26

Example: Pinning Model N Z ω Ω,β,h 1 + k=1 0<n 1 < <n k <N C k n 1 α 1 (n k n k 1 ) 1 α k k=1 ε xi k=1, contribution of one point + k=2, contribution of two points +... + k=m, contribution of m-points N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 21 / 26

Example: Pinning Model N Z ω Ω,β,h 1 + k=1 0<n 1 < <n k <N C k n 1 α 1 (n k n k 1 ) 1 α k k=1 ε xi Rescaling δ = 1 N and Gaussian Approximation Z ω Ω δ,β,h δ 1 1 + k=1 0<x 1 < <x k <1 δ (1 α)k C k x 1 α 1 (x k x k 1 ) 1 α k [ ] βδ 1 2 W(dxi ) + h δ 1 dx i k=1 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 21 / 26

Pinning Model and DPRE rescaling Pinning Model: γ = 1 α β δ = ˆβ δ α 1/2, h δ = ĥ δα must be α > 1/2 (Disorder relevance) ˆϕ (k) Ω (x 1,, x k ) = β δ = ˆβ δ 1/4, h δ = ĥ δ C k x 1 α (x 1 k x k 1 ) 1 α DPRE: γ = 2 ˆϕ (k) Ω ((x 1, t 1 ),, (x k, t k ))) = k e (x i x i 1 )2 /2(t i t i 1 ) i=1 2π(ti t i 1 ) N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 22 / 26

Universality of the free energy The continuum model captures the critical behavior of the system in the weak disorder regime (β, h 0): Naive approach ] 1 F ( ˆβ, ĥ) := lim [log Ω R d Ω E Ẑ W Ω =limδ 0 δ d Ω δ = Ω, ˆβ,ĥ 1 lim lim Ω R d δ 0 δ d Ω δ E [ ] log Z ω!!! Ω δ,β δ,h δ = lim lim 1 δ 0 Ω R d δ d Ω δ E [ ] log Z ω F(β δ, h δ ) Ω δ,β δ,h δ = lim. δ 0 δ d N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 23 / 26

Universality of the free energy The continuum model captures the critical behavior of the system in the weak disorder regime (β, h 0): Naive approach ] 1 F ( ˆβ, ĥ) := lim [log Ω R d Ω E Ẑ W Ω =limδ 0 δ d Ω δ = Ω, ˆβ,ĥ 1 lim lim Ω R d δ 0 δ d Ω δ E [ ] log Z ω!!! Ω δ,β δ,h δ = lim lim 1 δ 0 Ω R d δ d Ω δ E [ ] log Z ω F(β δ, h δ ) Ω δ,β δ,h δ = lim. δ 0 δ d Universal asymptotics F(β δ, h δ ) δ d ˆF(1, 1) as δ 0 N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 23 / 26

4 Proofs Continuum model Proof of our results

Recall the Results: Universal Critical Behavior Theorem (Caravenna, Toninelli, T., 2015) Universality of the Free Energy: F (α) F ( ˆβ, (α)( ˆβ ε α 1 2, ĥ ε α) ĥ) = lim ε 0 ε Theorem (Caravenna, Toninelli, T., 2015) h c (β) β 0 ĉ β 2α 2α 1, where ĉ universal constant depending on α and c. N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 24 / 26

Proof - Continuum Model τ := {n : S n = 0} ετ ε 0 Thm (C.S.Z.,15) (cond.) Pinning Model converges α (1/2, 1), β = ˆβε α 1 2, h = ĥε α : t 1 t 2 ετ β,h (d) ε 0 ˆτ ˆτ ˆβ,ĥ Continuum ingredients: - regenerative set (ˆτ) - White Noise (Cont. disorder) Problem: No Gibbs representation Wiener Chaos Expansion N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 25 / 26

Proof - Strategy: Coarse-Graining N = t ε. Consider 1 N E log Z N Convergence on each block Compare lim t lim ε & lim ε lim t Coarse-Graining of ετ and ˆτ 0 J 1= 1 J 2= 2 3 J 3= 4 J m= t 5 6 s t 1 s2 t 2 s3 t 3 sm t m t 1 t Partition function decomposition t s i t i (Ԑ) (Ԑ) Ԑ 0 s i t i Technical difficulty & c, disc Ԑ 0 Z (τ) Couple together convergence of ( si (ε), t i (ε) ) with Z c t/ε( a, b ) i c, cont Z i (τ) Z c 1(τ) Z c 2(τ) Z c 3(τ) Z c 4(τ) Z c,disc. i Z c,cont. i ( (τ/ε) = Z c t/ε si (ε), ) t i (ε) ( ) (ˆτ) = Z c si, t i t t Get: η > 0 ε 0 : ε < ε 0 F (α) ( ˆβ, ĥ η) ε 1 F (α)( ˆβ ε α 1 2, ĥ ε α) F (α) ( ˆβ, ĥ + η) N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 26 / 26

Proof - Strategy: Coarse-Graining N = t ε. Consider 1 N E log Z N Convergence on each block Compare lim t lim ε & lim ε lim t Coarse-Graining of ετ and ˆτ 0 J 1= 1 J 2= 2 3 J 3= 4 J m= t 5 6 s t 1 s2 t 2 s3 t 3 sm t m t 1 t Partition function decomposition t s i t i (Ԑ) (Ԑ) Ԑ 0 s i t i Technical difficulty & c, disc Ԑ 0 Z (τ) Couple together convergence of ( si (ε), t i (ε) ) with Z c t/ε( a, b ) i c, cont Z i (τ) Z c 1(τ) Z c 2(τ) Z c 3(τ) Z c 4(τ) Z c,disc. i Z c,cont. i ( (τ/ε) = Z c t/ε si (ε), ) t i (ε) ( ) (ˆτ) = Z c si, t i t t Get: η > 0 ε 0 : ε < ε 0 F (α) ( ˆβ, ĥ η) ε 1 F (α)( ˆβ ε α 1 2, ĥ ε α) F (α) ( ˆβ, ĥ + η) Similar to Copolymer Model (den Hollander & Bolthausen, 1997 and Caravenna & Giacomin, 2010) N. Torri (Université de Nantes) Random Polymers Paris, 17 Novembre 2015 26 / 26

Merci!

Merci!