Benchmark / Protocol

The benchmark protocol

A reviewer- and product-readable specification of what TargetSpace scores: the forecast unit, the sealed walk-forward loop, the scoring stack, and the reporting contract. This is the protocol a submission must satisfy.

Version  1.0 Status  pre-pilot protocol proposal Results  none reported no empirical results · no official submissions yet

1 · The forecast unit

Every scored item is one forecast unit: a sealed probabilistic prediction about a single future state of a single target, resolved later by a deterministic rule.

forecast unit  =  ( i , E≤t , q , A , r )   →   sealed probability vector over A
  • iInstance. The specific target and the specific occasion being scored.
  • E≤tEvidence up to t. All longitudinal evidence observable no later than the sealed time t. Strict walk-forward — no information from after t may enter the forecast.
  • qQuery. A question about a specific future state of the target.
  • AAnswer space. A discrete, enumerated set of possible outcomes.
  • rResolution time. A time r > t at which a deterministic, pre-registered observable rule assigns the realized outcome in A.
  • outputSealed forecast. The system emits a probability vector over A, sealed before the outcome exists.

Plain-language example: a note-taking app claims its memory helps its users. TargetSpace asks it to forecast — before the outcome exists — whether a user will complete, defer, cancel, or replace a recurring commitment. The forecast earns credit only if it beats population rates, beats that user's own routine, stays calibrated, and fails when scored against another user.

2 · The sealed walk-forward loop

One pass over a target produces one reporting row. The loop is strictly prospective: each forecast is sealed before its outcome exists, then scored against externally observable resolutions.

1
Observe up to tAssemble the evidence set E≤t under strict walk-forward. No future information may enter.
2
Seal the forecastEmit and freeze the probability vector over A before r; timestamp and hash it so it cannot be revised after the fact.
3
Resolve at rApply the deterministic, pre-registered observable rule to read the realized outcome in A.
4
Proper-scoreScore the sealed vector against the outcome with a proper scoring rule: log score in bits, plus Brier.
5
Compare against R1 and R2Skill in bits over the R1 population prior and over the R2 own-routine baseline, on identical sealed instances.
6
Calibration and permutationCheck calibration; re-score under matched target permutation, where target-specific skill must collapse.
gate
7
Evidence-tier liftAblate evidence tiers to measure the marginal contribution of each stream — measured, never assumed.
8
Emit the reporting rowRecord skill, calibration, permutation outcome, counts, horizon, domain, and result status.

The four-bar decision rule. A forecast earns target-specific credit only if it (1) beats the R1 population prior, (2) beats the R2 own-routine baseline, (3) stays calibrated, and (4) loses skill under matched target permutation. Miss any bar and the credit is not target-specific — the result reduces to generic prediction or routine replay.

3 · Scoring

Scoring is applied to the full sealed probability vector, not to cherry-picked accuracy. Skill is reported relative to two baselines, gated by calibration and permutation, and attributed across evidence tiers.

Log score (bits)

The primary proper scoring rule over the full vector. Higher is better; expressed in bits so skill differences read directly.

Brier score

A secondary proper score over the vector, reported alongside the log score.

Skill vs R1 (bits)

Bits gained over the population prior. Rules out looking good by predicting what usually happens; the entry condition.

Skill vs R2 (bits)

Bits gained over the target's own routine. Rules out routine replay: skill over R2 is skill the routine does not already contain — the headline number.

Calibration diagnostics

Calibration intercept and slope plus a reliability read. ECE is a coarse diagnostic at small n, not a hard gate.

Permutation specificity gate

Forecasts re-scored against the wrong target must lose their skill. If skill survives permutation, target-specific credit is withdrawn.

Evidence-tier ablation

The marginal lift of each evidence tier, measured by removing streams — what each stream buys, never assumed.

What a score does and does not certify. A high score certifies calibrated prospective predictive skill only — evidence that the system can maintain and use a target-specific predictive representation on sealed calibrated forecasts. It is not evidence of understanding, inner life, or causation, and it is never permission to act on a person.

4 · The hard inclusion rule

A target state is scored only if it has a pre-registered observable resolution rule; otherwise it is evidence, not a scored state. Attention, affect, and inferred goals are always evidence — never scored states. Self-report may be evidence, but it is never the outcome label when it is also a model input. Causal claims require intervention and are outside the default benchmark.

5 · Valid vs invalid target states

The inclusion rule, made concrete. Only states with a deterministic observable resolution rule are scored; everything else is admissible as evidence or excluded.

Candidate target stateCategoryVerdictWhy
P(no substantive reply to message m by r) Observable outcome Scored A deterministic observable rule reads the thread at r and assigns the outcome.
Commitment resolves as {complete, defer, cancel, replace} Observable outcome Scored Discrete answer space; each outcome is externally observable at r.
“Feels anxious” Affect Evidence only An affective state with no pre-registered observable resolution rule. Usable as evidence, never as a scored state.
“Attending to X right now” Attention Evidence only Attention is evidence, never a scored state.
“True goal is Y” Inferred goal Evidence only Inferred goals are evidence, never scored states.
“Attention causes the switch” Causal claim Out of scope A causal claim requires intervention and is outside the default benchmark.
A self-reported goal that is also a model input Contaminated label Invalid label Self-report may be evidence, but it is never the outcome label when it is also a model input.

6 · The reporting row

Each pass reports one row. A result is the row, not a single number — skill is meaningful only next to its baselines, gates, counts, and status.

  • Skill vs R1Bits gained over the population prior.
  • Skill vs R2Bits gained over the target's own routine — the headline.
  • Calibrationpass / warn / fail, or a diagnostic read at small n.
  • Permutationcollapses (target-specific) / survives (not target-specific).
  • Evidence liftMarginal bits attributable to each evidence tier.
  • Resolved nNumber of resolved sealed forecasts.
  • TargetsNumber of independent targets.
  • HorizonThe interval from sealed time t to resolution time r.
  • DomainThe target domain (for example, TS-Personal).
  • StatusResult status: synthetic / private / public.

A large forecast count from few targets is not a large independent sample. Resolved-forecast count and independent-target count are reported separately for exactly this reason.

Run the loop, or read the schemas

The Version 1.0 reference harness executes this loop on synthetic data as a smoke test and reference path — not a validated leaderboard result. The schema reference defines the on-disk formats a submission must produce.