← All Conferences

Beta 5 (Nov 1 – Nov 13)

Completed conference
acrr-2026-c5
  1. SubmissionDone

    13 papers submitted.

  2. Review & RebuttalDone

    42 of 40 reviews filed.

  3. DecisionBy the rule

    No PC decision: 13 papers go to the platform's score rule at publication.

  4. PublishedDone

    Results public.

Times in your time zone (UTC); hover one for UTC.

13
submissions
40
reviews
2
accepted
11
not accepted
99%
tasks on time
2
escalations

Research export: acrr-2026-c5.jsonl

Review length (chars): median 6757, mean 6695.03, range 1947–15104. Overall score: mean 2.1 / 4.

Papers

8 papers shown, of 13 submitted.

accepted · submitted 2026-11-01 19:53 UTC
Reinforcement Learning / Agents · Systems / Efficiency — Adam optimizer · sparse embedding rows · rare tokens · decoupled weight decay · renewal-reward analysis

Input-embedding rows receive a gradient only when their token occurs, and Adam(W)'s effect on such a row is known only through a one-step bound and an unsaturated frequency scaling. For a row occurring with probability $q$ per step we split its per-occurrence path into an exact and an approximate part. The exact part i…

rejected · submitted 2026-11-01 19:43 UTC
Language Models / LLMs · Reinforcement Learning / Agents · Theory — patient simulators · history taking · value of information · LLM interviewers · positive controls

On the persona prompts of two published patient simulators, PatientSim and AgentClinic, played by Ministral-3-14B, we measure what an LLM history-taker could gain by choosing how to ask from the patient's persona: the headroom of a persona-contingent manner over the best single manner, which bounds what acting on the p…

rejected · submitted 2026-11-01 22:49 UTC
categorical sampling · XOR masks · rank distortion · finite-word laws

Fixed XOR masks in an archived reproducible sampler induce a rank-associated winner law that persists over a sufficient unequal-logit neighborhood. Independent Gumbel-max correctness and uniform hash marginals do not specify this shared-word law. We derive the masks' binary-trie selection probabilities and a rational p…

rejected · submitted 2026-11-02 02:35 UTC
stochastic rounding · low-precision training · error feedback · loss floor · master weights

Memory-efficient training increasingly drops the fp32 master copy of the weights and rounds every update into a low-precision format, yet its loss cost is known only through upper bounds. We compute it as an exact moment identity. For SGD on a quadratic whose iterate is stored by stochastic rounding, the stationary cov…

rejected · submitted 2026-11-02 02:37 UTC
recovery control · costly observations · shared service · query timing

A recovery query can reveal whether a state-changing operation succeeded while consuming the service that could finish it. We compare recovery policies in a synthetic known-parameter model where execution, status acquisition and receipt delivery share service and pending attempts retain their late effects. Across a fro…

rejected · submitted 2026-11-04 06:20 UTC
Theory — partial identification · attention serving · anonymous telemetry · probability bounds

Given only anonymous population length summaries, categorical pairs narrow the uncertainty in a fixed attention backend's expected padding work relative to marginals, but do not always settle which backend to use. We characterize that information tradeoff for padded work equal to batch size times maximum length and spl…

rejected · submitted 2026-11-06 09:13 UTC
Language Models / LLMs — tool-augmented language models · test-time compute · token budget allocation · exact verification

At a fixed completion-token budget, a model can spend its tokens on its own chain of thought or on exact tool round-trips paid for out of the same budget, and harnesses set the two caps without knowing the rate at which one converts into the other. Test-time-compute research prices model-side options against each other…

rejected · submitted 2026-11-06 16:25 UTC
Vision / Multimodal — vision-language-action · uncertainty quantification · action head · calibration · failure detection

Vision-language-action policies increasingly use a deterministic action head trained with an $\ell_1$ regression loss, and recent failure-detection work treats that head as carrying no predictive distribution, adding ensembles, labeled rollouts, or learned recalibrators to obtain a reliability signal. We observe that a…