Last day. Blogging will be feeble.

**Force from non-equilibrium fluctuations**,* M. Kardar*. Fluctuation force: pressure of particles against the walls. As the temperature goes to zero the length scale becomes larger than the distance between the walls and one obtains Casimir forces [see “The theory of molecular attractive forces between solids”] Quantum effect. In classical systems you can get this kind of interactions if you have long-range correlations: the Goldstone mode, and when you have a nonequilibrium system. Rytov: fluctuational QED. Kardar considers an analogue of Rytov’s theory based on fluctuating hydrodynamics which is hydrodynamics with noise.

**Final session**. This was actually quite an interesting section. Lecomte’s talk on finite-time corrections to large deviations and dynamical phase transitions, an interesting talk on coalescent random walk (that is analytically solvable). Hartmann proposed a method to calculate work statistics using a computational trick based on generating a whole series of random numbers first, and then accepting events based on a modified “tilted” Metropolis rule, something that might provide an alternative to other algorithms like cloning. A nice large-deviation approach that they applied to an Ising model with 128^2 spins (quite a lot already). He also launched www.papercore.org.