Papers for

artificial intelligence developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Deterministic algorithm improves 3-SAT problem solving speed

Improved Deterministic Algorithms for $k$-SAT: Generalization of Chains and Recurrence Proof

Abstract: We present a deterministic algorithm for 3-SAT running in $1.32736^n$ time. The algorithm is built on a branching-and-chain framework augmented with a polynomial-time expansion procedure that joins short chains through additional clauses to form bounded-size blocks, which impose stronger constraints on the assignment space. After branching, the algorithm chooses between derandomized local search guided by the block structure and enumeration of block assignments together with selected port variables, reducing the remaining formula to 2-SAT. A weighted analysis of branching, partial-assignment, and local-search costs, together with inductive bounds for chains and block interactions, establishes the running-time guarantee. The framework is self-contained and also offers guidance for practical SAT solver design by identifying useful clause-overlap structures and suggesting when to switch between local-search and reduction-based strategies.

Thu 12 NovData Structures and Algorithms
The gist
Solving 3-SAT, a problem that asks if there is a way to assign true or false values to variables in a logical formula, can be very slow. The authors present a method that speeds this up deterministically by grouping clauses into small blocks and using a combination of search and reduction techniques. Their approach carefully balances different strategies and proves the improved speed mathematically. This method also offers ideas for building practical SAT solvers by showing how to choose when to switch strategies based on clause overlaps.
Open → 2611.16814v1

Looped reasoning methods solve problems faster with stable cycles

Looped Reasoning Finishes Earlier Than You Think

Abstract: Looped reasoning with dynamic depth can be viewed as relaxation toward solution attractors in a learned dynamical landscape. In practice, however, latent trajectories often oscillate, cycle, or drift rather than cleanly converge to stable fixed points. How these non-stationary dynamics contribute to reasoning and constrain convergence remains unclear. We show that slowly decaying oscillatory modes can impose their convergence rate on coupled latent dynamics, creating a bottleneck. Yet damping these modes indiscriminately can disrupt the reasoning computation they support. This tension motivates an architectural approach: introducing rotational degrees of freedom along which symmetry-generated rotational motion leaves the reasoning-relevant quotient dynamics unchanged by construction. We propose QRR (Quotient Recurrent Reasoner), a gauge-equivariant recurrent architecture with an exact per-token phase-rotation symmetry. Motion along these symmetry directions leaves the quotient representation unchanged and is dynamically decoupled from reasoning-relevant dynamics. Empirically, QRR achieves stable and faster convergence to solution roots in quotient space, without forming a dominant slow oscillatory bottleneck, while matching or surpassing recent recurrent reasoning models on Sudoku, Maze, and Mini-ARC.

Tue 13 OctMachine LearningArtificial Intelligence
The gist
Some computer reasoning methods take longer because their internal steps cycle or oscillate without settling down quickly. The authors show that these slow cycles can cause delays but can’t be removed entirely without hurting the reasoning process. They design a new method called QRR that separates the harmless oscillations from the reasoning steps, allowing the system to finish reasoning faster and more reliably. They tested this on puzzles like Sudoku and mazes, where QRR performed as well as or better than other recent systems.
Open → 2610.18174v1

Drosophila brain wiring supports learning human cognitive tasks

A Drosophila Whole-Connectome Network Can Learn Human-Designed Cognitive Tasks

Abstract: Can a biological wiring diagram serve as a useful computational substrate beyond the behaviors for which it evolved? We use the publicly released MaleCNS v1.0 connectome, reconstructed from a single adult male Drosophila specimen, as the fixed recurrent topology of an artificial network. We train separate models for bounded addition and for a controlled grounded relational language task built from a fixed 100-word lexicon. In both models, one scalar is learned per anatomical edge. The anatomical graph reaches 92.77% mean accuracy on held-out addition, compared with 67.93% for directed degree-preserving rewires. On the strict paired language endpoint, which matches original and order-reversed scenes to their corresponding descriptions, it reaches 61.59% across four fixed interfaces, compared with 44.17% for matched rewires. At the canonical interface, it ranks first in a fixed 21-graph comparison. On the matched 48-group intervention subset, shuffling task-defined sensory features reduces its score from 60.94% to 19.27%. Together, these results show that higher-order MaleCNS wiring provides a reusable inductive bias for bounded addition and grounded relational language.

Wed 7 OctMachine Learning
The gist
This paper shows that the detailed wiring of a fruit fly’s brain can be used as a starting point for machines to learn tasks designed by humans, like simple addition and understanding language relationships. The authors used a full brain map from one male fruit fly to build an artificial network and trained it on specific cognitive tests. Their models performed better than random rewired networks, suggesting the fly’s brain wiring contains helpful patterns for learning. This means biological brain structures might offer useful guidance for building smarter artificial systems.
Open → 2610.10014v1

Organizing cognitive schemas by information to improve problem solving reuse

Cognitive Schemas, Laws and Tasks

Abstract: This paper asks how explicit representations can support reusable cognitive schemas in knowledge-based problem solving. We develop a structural framework in which schemas are organized by the information and relations required for their use, rather than introduced as unrelated primitives. The framework also distinguishes context-dependent relations from more stable structures that can be reused across different representations. Tasks are described through the information available, the unknowns to be determined, and the constraints that admissible solutions must satisfy. This makes it possible to separate limitations of the representation from limitations of the solving procedure. In particular, we distinguish inconsistency, underdetermination, and contextual insufficiency, where the current representation lacks distinctions or relations required by the external task meaning. We also show formally when a reduction of representation preserves the task-relevant solution structure. The resulting task--schema interface offers a structured way to describe representational conditions relevant to problem solving. It supports the reuse and stabilization of derived knowledge while remaining independent of the particular mechanism used to generate candidate solutions. This may provide a useful component for future solver architectures that combine structured knowledge, verification, and learned proposal mechanisms.

Wed 7 OctArtificial Intelligence
The gist
Figuring out how our mental shortcuts or 'schemas' can be clearly represented to help solve problems better is a tricky challenge. The authors created a framework that organizes these schemas based on the information and relationships they need instead of just treating them as isolated concepts. This approach helps separate what limits our understanding from what limits how we solve problems and shows when simplifying representations won't lose important details. Their work provides a way to reuse knowledge across different tasks, which could help build smarter systems that combine known facts and learned strategies.
Open → 2610.09495v1

Student-conditioned updates improve teacher performance in on-policy distillation

On the Off-Policy Teacher in On-Policy Distillation

Abstract: On-policy distillation (OPD) has recently emerged as a promising post-training paradigm in which the student learns from trajectories generated by its own policy under dense teacher supervision. However, OPD introduces a fundamental asymmetry: although the sampled trajectories are on-policy for the student, they are off-policy for the teacher. The teacher is typically optimized to continue from prefixes generated by its own policy, but during OPD it must instead supervise prefixes generated by the student. Empirically, we find that its continuation performance degrades as these prefixes grow longer. To address this issue, we propose Student-COnditioned Updates of the Teacher (SCOUT), a co-training framework that adapts the teacher to student-generated prefixes. Alongside standard OPD updates, SCOUT periodically optimizes the teacher's conditional ability using reinforcement learning with verifiable rewards, where the teacher generates continuations from student prefixes and learns from outcome rewards. Controlled experiments show that SCOUT improves the teacher's ability to continue from student-generated prefixes, supporting the intended mechanism of student-conditioned teacher adaptation. Across multiple teacher--student configurations, model scales, and reasoning domains, SCOUT also consistently improves the effectiveness of on-policy distillation.

Tue 29 SeptMachine LearningArtificial IntelligenceComputation and Language
The gist
This paper addresses a challenge when teaching AI models called on-policy distillation. In this process, the student model learns from its own experiences but is guided by a teacher model that was trained on different behaviors. The authors find that the teacher struggles to guide the student well as the student’s behavior differs more from what it originally learned. They propose a method called SCOUT, where the teacher is periodically retrained to better handle the student’s behavior. This approach helps the teacher give better guidance, improving the student’s learning across different setups.
Open → 2609.38360v1

Randomized models often outperform worm brain networks in tasks

Benchmarking the Connectomes of Caenorhabditis elegans within the Reservoir Computing Framework

Abstract: The aim of this work is to examine the connectomes of Caenorhabditis elegans through a computational lens using the reservoir computing framework. Connectomes are mappings of biological neural networks; C. elegans is the first organism for which physical connectomes covering the whole nervous system have been published. The connectomes of C. elegans used in this paper have been derived at different ages of the organism and are based on three different ways of measuring inter-cellular connections. They have, with minimal preprocessing, been implemented as reservoirs in the form of echo state networks, which are recurrent neural networks. In reservoir computing, the reservoir itself is not trained, rather the output of the reservoir is passed to a comparatively small read-out module in which training takes place. Training and testing is conducted in different neuro-inspired tasks, with the aim of using these tasks as a benchmark for the connectomes. This process has been repeated with different configurations of the reservoir and equally sized but randomized null models have been used for comparison. The results show that the biological wiring and a bio-informed configuration of input and output nodes of the reservoirs do not necessarily lead to better performance. Contrarily, the randomized null models are often outperforming the original connectomes on the chosen benchmarks. At the same time it becomes clear that the results depend a lot on the configuration of the reservoir and the way the connectome has been derived from the organism. Connectomes from different ages may produce varying outcome, without a clear trend becoming visible.

Thu 24 SeptMachine Learning
The gist
This paper looks at the nerve connections of a tiny worm, C. elegans, to see if its brain wiring is better than random setups at certain tasks. The researchers tested worm brain maps at different ages using computer models called reservoir computing. Surprisingly, random networks often did better than the actual worm brain connections in the tests. They also found that performance depends a lot on how the model is set up and which brain map is used.
Open → 2609.30508v1

Brain size grows first in stable environments before changing ones

Big Brains and Changing Environments: Cause or Consequence?

Abstract: Large brains are metabolically costly, and associations with changing environments do not imply they evolved there, as the Cognitive Buffer Hypothesis (CBH) would suggest. They may instead evolve in stable conditions and later facilitate colonization of changing environments. Using neuro-evolution in an artificial seasonal foraging task, we compared agents evolving exclusively in changing environments to agents first evolved in static environments before transitioning. Results show that larger neural networks in dynamic environments arise mainly from prior static evolution, achieving superior performance under unpredictable changes. Our results challenge strict CBH predictions, provide agent-based (computational) support for a colonization-based account and highlight the role of evolutionary history in brain size evolution.

Mon 14 SeptNeural and Evolutionary ComputingArtificial Intelligence
The gist
Bigger brains cost a lot of energy, so scientists want to know why animals evolved them. This study shows that animals might develop large brains in stable environments first, not just in changing ones as some ideas suggest. Later, these bigger brains help animals survive better when their surroundings become unpredictable. The authors ran computer simulations where virtual agents evolved bigger brains in static settings before facing changing conditions, and those with large brains did better. This challenges previous beliefs and highlights that the history of evolution matters in brain size changes.
Open → 2609.15569v1