1: \begin{abstract}

2: This paper elaborates control strategies to prevent clustering effects in opinion formation models. 

3: This is the exact opposite of numerous situations encountered in the literature where, on the contrary, one seeks controls promoting consensus.

4: % Interestingly, maximizing the classical variance does not prevent clustering.

5:  In order to promote declustering, instead of using the classical variance that does not capture well the phenomenon of dispersion, we introduce an entropy-type functional that is adapted to measuring pairwise distances between agents.

6: We then focus on a Hegselmann-Krause-type system and design declustering sparse controls both in finite-dimensional and kinetic models. 

7: We provide general conditions

8: characterizing whether clustering can be avoided as function

9: of the initial data. Such results include the description of black

10: holes (where complete collapse to consensus is not avoidable),

11: safety zones (where the control can keep the system far from

12: clustering), basins of attraction (attractive zones around the

13: clustering set) and collapse prevention (when convergence to the clustering set can be avoided).

14: 

15: \paragraph{Keywords:} Active particles, collective behavior, control, kinetic model, declustering, black swan.

16: 

17: \paragraph{AMS Subject Classifications:} 93C15, 93C20, 91D10, 35B36, 34H05.

18: \end{abstract}

19: