Proof of Theorem 4.1, p. 17 — if , and , then
ProvedEngagementOpt.Misalign.thm_4_1_step_vbehavioral-economicsengagementp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1recommender-systems
This is the second step of the proof of Theorem 4.1.
Let and let and be two points of , with
Suppose that
- ;
- ;
- ;
- .
Then
Together with the first step, this shows that when the engagement maximizer is not more moreish, it has a longer span but a lower value.
Formalization Note Hypothesis 4 is the constraint of (4) satisfied by the engagement maximizer; the printed step marked "" uses it silently (it gives ), and the statement is false without it. is added: the paper fixes (p. 5).
Preamble
import Mathlib import Definitions.Def_EngagementOpt_Misalign_Model
Formal statement
namespace EngagementOpt.Misalign
/-- Proof of Theorem 4.1, p. 17: for `ω_S, ω_T ∈ Ω` and `W > 0`, if `p_T ≤ p_S`, `q_T > q_S`,
`g_S(ω_S) > g_S(ω_T)` and `g_S(ω_T) ≥ 0` (the constraint of (4), used implicitly in the step
marked "(q_T > q_S)"), then `v̄_T < v̄_S`. -/
theorem thm_4_1_step_v (W : ℝ) (hW : 0 < W) (ωS ωT : Params) (hS : InOmega ωS)
(hT : InOmega ωT) (hp : ωT.p ≤ ωS.p) (hq : ωS.q < ωT.q) (hgS : gS W ωT < gS W ωS)
(hpart : 0 ≤ gS W ωT) :
ωT.v < ωS.v := by sorry
end EngagementOpt.Misalign
Source
Kleinberg, Mullainathan, Raghavan, The Challenge of Understanding What Users Want, arXiv:2202.11776v3, p. 17, proof of Theorem 4.1 (second display)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.