Does anyone know where I can find an explicit formula to compute the level 2 or level 3 characters of affine $su(2)$? I have found several sources that give a formula to compute the level 1 characters in terms of theta functions, but I cannot find anywhere in the literature nice formulas for level 2 or level 3. Thanks

The books Affine Lie algebras, weight multiplicities, and branching rules by Kass, Moody, Patera, and Slansky contain a lot of such explicit computations (I don't have them right now in front of me, so I don't know if this answers your specific question). Also, Volume 2 of that series of two books contains a lot of tables that might be useful for what you're doing. At any rate, I highly recommend those books: they are very readable and contain a lot of information. That's where I learned that subject. 


I would suggest "Fock Representations of the Affine Lie Algebra $A^{(1)}_1$ by Wakimoto. 


By "level 2 or level 3 characters of affine $su(2)$" I hope you meant "characters of level 2 or 3 integrable highest weight $A_1^{(1)}$modules". Also, I assume you don't want principally specialized characters. Those can be easily computed by using the socalled "numerator formula". See this paper for more details in the $A_1^{(1)}$ case. For example, level three (principally specialized) characters are known to be RogersRamanujan series multiplied with something like $\prod_{n=1}^ \infty (1q^{2n1})^{1}$. Level 2 modules can be realized via fermions, so not surprisingly their characters admit nice form as in the level one case. Here's what you get directly from WeylKac: NeveuSchwarz sector: $$\chi_{L(2 \Lambda_0)} \pm e^{\delta/2} \chi_{L(2 \Lambda_1)}=e^{2 \Lambda_0} \left( \prod_{n=1}^\infty (1 \pm e^{\delta(n1/2)})(1e^{n \delta})^{1} \right) \sum_{n \in \mathbb{Z}} (\pm 1)^n e^{n \alpha_1} e^{\delta n^2/2}$$ Ramond sector: $$\chi_{L(\Lambda_0+\Lambda_1)}=e^{\Lambda_0+\Lambda_1} \left( \prod_{n=1}^\infty (1 + e^{\delta n})(1e^{n \delta})^{1} \right) \sum_{n \in \mathbb{Z}} e^{n \alpha_1} e^{\delta (n^2+n)/2}$$ I guess you can write down (more complicated) level 3 characters. But if I remember correctly level 3 modules have no known bosonic/fermionic constructions, so I doubt they will look as nice as those of level 1 and 2. 


Affine Lie algebra : Let $\mathfrak{g} = \mathfrak{su}_{2}$, $L\mathfrak{g}$ its loop algebra and $\mathcal{L}\mathfrak{g} = L\mathfrak{g} \oplus \mathbb{C}\mathcal{L}$ its affine Lie algebra (the central extension) :
Character at level $\ell$ and spin $i$ : Definition : Let $m \in \mathbb{N}^{\star} $, $n \in \mathbb{Z} $, $t, z \in \mathbb{C}$ with $\Vert t \Vert < 1$. References :


