← Research map
Technical noteObserve

Finite-memory adaptive tracking

How much acquired information survives compression when the target itself can change?

Read Online Or start with the explanation below.

Why this question matters

An adaptive system can learn an answer, discard it through compression, or change the target through action. Counting memory variability alone conflates these events. The note separates their contributions before asking which finite memory and sensing policies are useful.

Setup and results

The accounting identity states exactly when compression loses no target information. A sharp entropy bound forces loss under a finite memory budget. Two-bit and finite-horizon examples separate retained information from later query performance. Exact selective sensing can beat fixed schedules in a certified parameter box; a six-step counterexample defeats a shorter-horizon pattern.

Limits and negative results

The finite-horizon recurrence specializes established control methods. No new general information theory, global optimality of the periodic controller, or cognitive implementation is established. A clock supplied for free changes the storage count when compiled into a controller.

Reading guide

Start with the accounting identity and worked counterexamples, then read the tracking and uniform-comparison extensions. Compare the finite synthesis recurrence with the explicit sensing and memory prices before interpreting the plots.

Place in the research programme

This extends the acquisition family by pricing persistent memory and tracking a changing target. Restricted noisy experiments studies which supplied sensing interfaces can acquire a target; adaptive bit acquisition separately counts streaming state costs.

Edition and review status

Brian Theory is the author; this is an AI-assisted research draft. Documented specialist and hostile AI review rounds informed development, but not every round was blind, independently provisioned or cover-to-cover on the final assembly. These checks are not external human peer review or journal acceptance. Targeted human review and historical novelty assessment remain pending.

Sources and reproducibility

Read the complete manuscript, open its page-faithful PDF reader, or download the exact PDF. Exact LaTeX and figure dependencies contain the manuscript inputs and a checksum manifest. This source-only package does not contain the computational verification programs; their public release and independent execution are not certified here. Read the manuscript’s reproducibility sections for the scope of its finite checks.

PDF SHA-256: 9a1b126646e2515f7114be9beb529e481b723718d2950867bf901e1806b8a0bd. Main LaTeX SHA-256: a760946faf45049ae171769c753f16a43b8e334ac3ddde88eeada9086c22a49a.

Related manuscript

Acquisition, Forgetting, and Target Change: Information Accounting for Adaptive Systems

AI-assisted research draft; internal AI checks documented. External human review and novelty assessment remain pending.

Reading scope: Finite targets, declared pre-observation records, priced sensing and persistent finite memory. The periodic controller is an achievable upper bound, not a global optimum.

Site reading copy 1 from 13 November 2026 manuscript with localized corrections and approved author credit; exact edition hashes linked below.

Technical material

Exact LaTeX source and figure dependencies

See this manuscript's index entry
Phase: ObserveObserveManuscript: Acquisition, Forgetting, and Target Change: Information Accounting for Adaptive SystemsAcquisition, Forgetting,…Project: Adaptive bit acquisition and memoryAdaptive bit acquisition…Project: Restricted noisy experimentsRestricted noisy experime…Finite-memory adaptive tracking
  • Phases
  • Manuscripts
  • Projects

Conceptual and navigational links, not proof dependencies.