The workshop will last four full days.

To meet the identified purposes, the program will comprise:
- First two days: A few « perspective talks » where a participant is invited to present the state-of-the-art of a specific class of method, including implementation details, current theoretical, mathematical and implementation challenges, and possible future directions. A time slot of 1h30 will be allocated, but material for no more than a ~35-40 minutes talk will be prepared, in order to allow comments, questions and additions from other participants during the course of the presentation ;
- First two days: a few short talks that expose other facets of the class of methods presented in the « perspective talks » ;
- Last two days: only « framed discussions »: time slots of 1h30 dedicated to a precise topic to be addressed. It is envisage that two to three participants will introduce a precise point, question or challenge in no more than 10 minutes, to be followed by extensive discussions that hopefully lead to us starting to working collectively on the particular challenges that have been highlighted;
- All the way through the meeting: extended coffee breaks and lunch times, together with common dinners, in order to facilitate the continuation of lively discussions in small groups in a more informal and flexible way.

Small offices will be available for participants to pursue informal discussions in smaller groups (during lunch time, or before dinner time).

The workshop will be held either in the "Université Pierre et Marie Curie" campus, or in the « Institut Henri Poincaré » in Paris (http://www.ihp.fr/en), both located at the heart of the 5th arrondissement in Paris, and that could make available conference rooms of suitable size and facilities, as well as offices for the additional small group discussions. The coffee breaks and lunches (buffets) will take place in the same building, facilitating continuous and efficient interactions between participants.

- The five « perspective talks » will be:
1- A « mathematical » introduction to Valence Bond theory
2- Second quantization techniques and reduced density matrix formalism for nonorthogonal wave function methods.
3- Nonorthogonal Valence Bond ab initio methods and implementation.
4- Nonorthogonal CI approaches.
5- Geminal-based wave function methods with non-orthogonal orbitals.

The possible « challenges » for non-orthogonal orbital based methods that may be addressed during the four « framed discussions » could be (the final list will be decided after interactions amongst the participants):
1- Efficient and cheap inclusion of dynamical correlation.
2-,New algorithms for nonorthogonal orbitals.
3- Strategies for excited states.
4- Complex systems (metals, extended systems)
5- Efficient implementation, and massively parallel computer architectures.