discreteness_forcing_principle
plain-language theorem explainer
The discreteness forcing principle asserts that the cost J(x) = ½(x + x⁻¹) - 1 with unique minimum at x=1 and second derivative 1 in log coordinates implies stable existence requires discrete configuration space. Researchers deriving ontology from the Recognition Composition Law cite it as the bridge from cost landscape to discreteness. The proof assembles four lemmas on nonnegativity, uniqueness, curvature, and non-isolation into a single conjunction.
Claim. $J(x) = ½(x + x^{-1}) - 1$ for $x > 0$ satisfies $J(x) ≥ 0$, $J(x) = 0$ if and only if $x = 1$, the second derivative of $J(e^t)$ at $t=0$ equals 1, and every zero of $J$ is a limit point of the domain.
background
The DiscretenessForcing module shows that the cost landscape forces discrete structure. J_log(t) := cosh(t) - 1 is the cost in logarithmic coordinates, a convex bowl with minimum at t=0. The defect functional equals J(x) for x>0. Upstream results establish defect_nonneg (J ≥0), defect_zero_iff_one (unique zero at 1), and J_log_second_deriv_at_zero (curvature 1 at minimum). The module_doc states that continuous spaces permit infinitesimal perturbations with infinitesimal cost, precluding stable minima.
proof idea
Term-mode proof packages the four conjuncts directly: defect_nonneg, defect_zero_iff_one, J_log_second_deriv_at_zero, and an explicit construction that substitutes x=1 via defect_zero_iff_one then picks the point 1 + ε/2 to witness non-isolation.
why it matters
This theorem occupies Level 2 in the forcing chain (Composition law → J unique → Discreteness forced). It is cited by zero_param_forces_scale_free as the zero-parameter bridge to scale-free structure. The doc_comment identifies it as the step from J-uniqueness to the requirement of discrete configuration space for RSExists, preceding the phi-ladder and D=3.
Switch to Lean above to see the machine-checked source, dependencies, and usage graph.
papers checked against this theorem (showing 8 of 8)
-
Spann3R predicts global pointmaps via external spatial memory
"Spann3R shows competitive performance and generalization ability on various unseen datasets and can process ordered image collections in real time."
-
Frontier models scheme to disable oversight and exfiltrate weights
"We study whether models have the capability to scheme in pursuit of a goal that we provide in-context and instruct the model to strongly follow."
-
Looped models match 12B LLMs with 1.4B params
"Through controlled experiments, we show this advantage stems not from increased knowledge capacity, but from superior knowledge manipulation capabilities"
-
Constant-memory agent beats larger model on 16-objective tasks
"At each turn, MEM1 updates a compact shared internal state... pruning the agent's context to retain only the most recent <IS>"
-
Pixel ops lift 7B VLM to 84% on visual reasoning tests
"the model’s initially imbalanced competence and its reluctance to adopt the newly introduced pixel-space operations... curiosity-driven reward scheme"
-
LLM agents follow malicious instructions without jailbreaks
"We evaluate a range of leading LLMs, and find (1) leading LLMs are surprisingly compliant with malicious agent requests without jailbreaking, (2) simple universal jailbreak templates can be adapted to effectively jailbreak agents, and (3) these jailbreaks enable coherent and malicious multi-step agent behavior and retain model capabilities."
-
Persona vectors track and steer AI personality shifts
"We find that both intended and unintended personality changes after finetuning are strongly correlated with shifts along the relevant persona vectors. These shifts can be mitigated through post-hoc intervention, or avoided in the first place with a new preventative steering method."
-
Intelligence is skill-acquisition efficiency, not task performance
"skill is heavily modulated by prior knowledge and experience: unlimited priors or unlimited training data allow experimenters to buy arbitrary levels of skills"