14:32:58	 From Yogeshwaran D : https://www.isibang.ac.in/~d.yogesh/BPS/
14:33:40	 From Yogeshwaran D : Mathew’s slides are can be found on this page.
14:34:09	 From Parthanil Roy : And Siva’s will be uploaded later, right?
14:34:18	 From Yogeshwaran D : yes. After the talk.
14:38:42	 From Arvind Ayyer : This is assuming X_0 = 0?
14:39:17	 From Parthanil Roy : Can Siva scroll down a bit?
14:39:33	 From Mathew Joseph : yes to Arvind
14:39:52	 From Arvind Ayyer : Thanks
14:40:07	 From Parthanil Roy : Thanks
14:43:21	 From Parthanil Roy : This can be done for any self-similar process, right?
14:43:40	 From Mathew Joseph : Yes, that’s right.
14:43:46	 From kram : So if we know the torsion function (u such that laplacian u = -2 with u = 0 on boundary), then bounds for the exit time are immediate
14:44:10	 From Parthanil Roy : Thanks Mathew.
14:44:15	 From Yogeshwaran D : Does Mogulskii’s result have a path wise version too ?
14:44:46	 From Parthanil Roy : Varadhan-Mogulskii, right?
14:51:02	 From Abhay Gopal Bhatt : sup is over both s, t < = 1?
14:51:17	 From Mathew Joseph : yes
14:51:28	 From Yogeshwaran D : No lower bound for L_2 norm ?
14:51:30	 From Abhay Gopal Bhatt : ok. Thanks
14:52:10	 From Mathew Joseph : Yogesh: I don’t think he obtained a lower bound
14:52:51	 From kram : yes, thanks yogesh
14:57:06	 From kram : Matthew, the L2 norm is less than the sup norm, so doesn't this automatically give lower bound for small ball for ||B||_2?
14:57:14	 From Parthanil Roy : Is it Brownian sheet or BS conditioned to be something?
14:57:35	 From Srikanth Iyer : In the application B(1,1) is zero.
14:57:52	 From Parthanil Roy : Yes, so BS Bridge
15:00:23	 From Parthanil Roy : Siva, would you mind switching off your video? breaking up a bit from time to time.
15:01:04	 From Parthanil Roy : thanks
15:13:44	 From Yogeshwaran D : Shouldn’t it be P(A_0) \geq C for some constant ?
15:35:00	 From Parthanil Roy : can you please explain what is meant by periodic boundary conditions?
15:35:25	 From Parthanil Roy : Thanks
15:35:28	 From Siva Athreya : Boundary values match
15:35:44	 From Parthanil Roy : Got it, thanks.
15:56:05	 From Koushik : Mathew and Siva - Do you have an inequality like 
15:56:11	 From Siva Athreya : ?
15:56:37	 From Siva Athreya : the part after like did not come ?
15:57:58	 From Koushik : sup |u|\leq C_n sup|p_{i.j}| by some kind of net argument (using Lipschitz nature or anyhting else?)
15:58:50	 From Koushik : sorry i meant sup |u(p_{i, j})|
15:58:54	 From Siva Athreya : u has a stochastic integral
15:59:17	 From Siva Athreya : in probability we can compare
15:59:23	 From Siva Athreya : but almost sure will be hard to do
15:59:38	 From Koushik : I see, ok. Thanks
16:09:00	 From Parthanil Roy : proved by a retd German professor, right?
16:09:17	 From Yogeshwaran D : yes.
16:10:45	 From Yogeshwaran D : https://www.quantamagazine.org/statistician-proves-gaussian-correlation-inequality-20170328/
16:10:48	 From Koushik : sorry, can you explain again how you apply correlation inequality to estimate the underlined quantity?
16:10:58	 From Siva Athreya : He is doing it.
16:11:08	 From Balarka Sen : Symmetric means radially symmetric around the origin, or?
16:11:57	 From Siva Athreya : Koushik: Split the rectangle into smaller rectangles. Use noise part in each rectangle to get collection of Gaussian random variables
16:14:15	 From Koushik : Ah ok , got it
16:15:15	 From Yogeshwaran D : @Balarka, I guess centrally symmetric i.e., x \in K iff -x \in K.
16:15:46	 From Balarka Sen : Ah alright.

16:23:35	 From Sarvesh Iyer : Can you elaborate a little more on the approximation argument for general \sigma functions? (Is it standard?)

16:27:33	 From Balarka Sen : But what if K and L are disjoint centrally symmetric subsets of R^2? K = (very small disk at (1, 1)) U -(very small disk at (1, 1)) and L the same but with very small disk at (2, 2)?