Raphael Avalos

[NEW JOB]

member of technical staff @cohere

[NEW PAPER]

shiq paper (value base alg for llms) was presented at neurips arxiv

> i post-train llms with rl @cohere

> research, mostly rl, planning & llms

> education = {
phd candidate @ailab_vub @fwo,
meng cs @telecom,
msc applied math @mva_ens,
msc data science @eurecom
}

> tools = [linux, python, jax, vim]

> workshop_organization = [EWRL23, ALA24, ALA25]

> full_resume

selected_papers

[ICLR24] The Wasserstein Believer: Learning Belief Updates for Partially Observable Environments through Reliable Latent Space Models
Raphael Avalos*, Florent Delgrange*, Ann Nowe, Guillermo Perez, Diederik M Roijers
[ pdf ] [ bibtex ]

[RLC24] Online Planning in POMDPs with State-Requests
Raphael Avalos, Eugenio Bargiacchi, Ann Nowé, Diederik M Roijers, Frans Oliehoek
[ pdf ] [ bibtex ]

[TMLR23] Local Advantage Networks for Multi-Agent Reinforcement Learning in Dec-POMDPs
Raphael Avalos, Mathieu Reymond, Ann Nowe, Diederik M Roijers
[ pdf ] [ bibtex ]

news

[2025-09-28]

back @cohere to finetune more llms

[2025-05-20]

shiq paper (value base alg for llms) on arxiv

[2025-04-01]

command a technical report published arxiv

contact_me

firstname@lastname.fr