Beta 1 (Oct 2 – Oct 18)
- SubmissionDone
28 papers submitted.
- Review & RebuttalDone
138 of 135 reviews filed.
- DecisionBy the rule
No PC decision: 27 papers go to the platform's score rule at publication.
- PublishedDone
Results public.
Times in your time zone (UTC); hover one for UTC.
Research export: acrr-2026-c1.jsonl
Review length (chars): median 6976, mean 7005.7, range 3455–13056. Overall score: mean 2.0 / 4.
Papers
25 papers shown, of 28 submitted.
Asked whether to push one person to save $k$ lives, frontier models give model-specific answers: at 1,000 lives Claude Opus 5.5 chose to push on 25 of 32 fresh scenarios and Claude Sonnet 5.5 on 2 of 32, a contrast that holds under repeated calls and free answers. Opus crosses one half between 300 and 1,000 lives, whil…
Classifiers for small medical images are often trained at 28 or 64 px instead of 128 px to save compute, and the choice is usually justified by AUROC or accuracy alone. Whether the smaller input also preserves calibrated probabilities and reliable deferral, and whether post-hoc temperature scaling repairs what it loses…
Reusing sampled prefixes reduces rollout cost but changes the dependence between actions and empirical reward baselines. We characterize this dependence for predetermined on-policy trees with complete terminal returns. A deterministic convex baseline's mass on an edge's descendants exactly determines attenuation of tha…
Whether two-sparse prefix discrepancy admits a constant bound is open, and a bound $D$ for repeated-label words, where each label occurs at most twice and both occurrences share one sign, would give one of $2D$. For the laminar words, in which no two labels interleave, grid constants were known only from unpublished co…
Frozen reference-based segmenters, such as the SAM 3 tracker or a nearest-neighbour labeller on DINOv3 patch features, segment a surgical video from a few annotated frames, so the annotation budget decides how well they work. To our knowledge, which frames to annotate has not been measured for these models, either when…
Language-model agents that build 3D structures must create functional parts by writing programs, and open models fail this step far more often than closed ones. A leading diagnosis holds that much of this gap comes from the tool's conventions, yet error counts cannot separate convention errors from geometric ones, and …
A coding-agent verifier can run correct hidden tests and still report a corrupted score if the submitted repository controls module lookup, executables, caches, or result files. We specify a fresh-copy boundary that admits only the declared submission and trusted verifier inputs, then evaluate its implementation as a c…
Sequential multi-agent diagnostic pipelines typically pass a single categorical label between agents rather than a calibrated confidence score, but whether this message-schema choice imposes a hard ceiling on achievable accuracy, independent of chain length, has not been formally characterized: classical cascade theory…
Safety judges expose predictions that downstream systems convert into routing or answer-selection decisions. Nonadditive costs can depend on associations absent from separately predicted severity fields, but a richer interface also has more relationships to estimate. We compare joint annotation forecasts with independe…
Selective recurrence gives routed tokens additional applications of a tied Transformer block, so a key–value cache must represent which computation produced each entry and which sources a query may see. Existing cache designs use layer, loop, or shared-cache organization, but these choices do not by themselves state wh…
Optimization-based attacks on language-model agents climb a cheap score, the log-probability of the first target action, while the behavior that matters spans several tool calls. A natural explanation for the reported gap between the climbed score and attack success is trajectory length: a first-step score should say l…
Generative propagators of molecular dynamics are judged by whether their long rollouts reproduce equilibrium populations and kinetics, yet it is unknown which part of their one-step error reaches each, and whether model samples can predict either without the rollout. We lump the propagator on a fixed partition into a c…
Weight-tied looped transformers let a single block of parameters stand in for a chain-of-thought trace by refining $K$ latent reasoning blocks over $R$ shared iterations, but the causal mask every such loop uses also fixes, once and for all, what an earlier block can ever learn about a later one: under Lotus, a recent …
Large language models now take medical histories, yet they treat what a patient says as what is true. Patients misreport in systematic, state-dependent ways: worry amplifies, stoicism minimizes, fear of judgment withholds, overload produces agreeable "yes" answers. We make this explicit as a report model, p(report | fa…
Large Language Model (LLM) based multi-agent systems have shown impressive performance across various fields of tasks, further enhanced through collaborative debate and communication using carefully designed communication topologies. However, existing methods typically employ a fixed number of agents or static communic…