Jack A. Thorne
Abstract:We define a reduction covariant for the representations a la Vinberg associated to stably graded Lie algebras. We then give an analogue of the LLL algorithm for the odd split special orthogonal group and show how this can be combined with our theory to effectively reduce the coefficients of vectors in a representation connected to 2-descent for odd hyperelliptic curves.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is to define a reduction covariant for the representations of stably graded Lie algebras and provide a method similar to the LLL algorithm for effectively reducing the vector coefficients related to the 2 - descent of odd hyperelliptic curves. Specifically:
1. **Define the reduction covariant**: The paper first defines the reduction covariant \(R: V_s(\mathbb{R})\to X_G\) in the representation of stably graded Lie algebras, where \(X_G\) is the symmetric space of the group \(G\). This step is achieved by introducing the concept of Cartan involution.
2. **Reduction algorithm**: Next, the paper proposes a reduction algorithm for the odd split special orthogonal group \(SO(2g + 1)\), which is based on the idea of the LLL algorithm and can move elements into the Siegel set. This algorithm is especially suitable for studying the 2 - Selmer group of odd hyperelliptic curves.
### Formula summary
1. **Definition of Cartan involution**:
- For the Cartan involution \(\theta\) on the semisimple Lie algebra \(\mathfrak{h}\), it satisfies one of the following equivalent conditions:
1. The symmetric bilinear form \(B_\theta(X, Y)=-B(X,\theta(Y))\) is positive definite.
2. The \(\mathbb{R}\)-vector subspace \(\{X\in\mathfrak{h}_\mathbb{C}\mid\theta(X) = X\}\) is a real form of \(\mathfrak{h}_\mathbb{C}\) and is compact (i.e., its Killing form is negative definite).
2. **Definition of reduction covariant**:
- For any stable vector \(T\in V_s(\mathbb{R})\), there exists a unique Cartan involution \(\theta_T\) such that \([T,\theta_T(T)] = 0\) and \(\theta_T(\mathfrak{h}_i)\subset\mathfrak{h}_{-i}\) for each \(i\in\mathbb{Z}/m\mathbb{Z}\).
3. **Mapping of reduction covariant**:
- The mapping \(R = R_h: V_s(\mathbb{R})\to X_G\) is defined as \(T\mapsto\theta_T|_G\), and this mapping is \(G(\mathbb{R})\)-equivariant.
4. **Calculation of reduction inner product**:
- Given \(T\in V_s(\mathbb{R})\), there exists a unique inner product \(H = H_T\) that is compatible with \(J\) such that \(T\) is normal with respect to \(H\), that is, \(T\) commutes with its \(H\)-adjoint \(H^TTH^{-1}\). At this time, \(R(T)=\theta_{H_T}\).
5. **Specific steps of the reduction algorithm**:
- Use the Gram - Schmidt orthogonalization process to calculate the matrices \(n_H\) and \(a_H\), and adjust these matrices according to specific conditions to ensure that the final result satisfies the definition of reduction.
### Application example
The paper shows through a specific example how to apply the above theory to the 2 - descent problem of odd hyperelliptic curves. Specifically, consider the case of \(g = 3\), and the corresponding Lie algebra \(\mathfrak{sl}_7\) and special orthogonal group \(SO(J)\). Given an integral point \(P\), construct a matrix \(T\) and use the reduction covariant and reduction algorithm on it.