Position: We Need An Algorithmic Understanding of Generative AI

• URL: http://arxiv.org/abs/2507.07544v1
• Date: Thu, 10 Jul 2025 08:38:47 GMT
• Title: Position: We Need An Algorithmic Understanding of Generative AI
• Authors: Oliver Eberle, Thomas McGee, Hamza Giaffar, Taylor Webb, Ida Momennejad,
• Abstract summary: This position paper proposes AlgEval: a framework for systematic research into the algorithms that LLMs learn and use.<n>AlgEval aims to uncover algorithmic primitives, reflected in latent representations, attention, and inference-time compute, and their algorithmic composition to solve task-specific problems.
• Score: 7.425924654036041
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: What algorithms do LLMs actually learn and use to solve problems? Studies addressing this question are sparse, as research priorities are focused on improving performance through scale, leaving a theoretical and empirical gap in understanding emergent algorithms. This position paper proposes AlgEval: a framework for systematic research into the algorithms that LLMs learn and use. AlgEval aims to uncover algorithmic primitives, reflected in latent representations, attention, and inference-time compute, and their algorithmic composition to solve task-specific problems. We highlight potential methodological paths and a case study toward this goal, focusing on emergent search algorithms. Our case study illustrates both the formation of top-down hypotheses about candidate algorithms, and bottom-up tests of these hypotheses via circuit-level analysis of attention patterns and hidden states. The rigorous, systematic evaluation of how LLMs actually solve tasks provides an alternative to resource-intensive scaling, reorienting the field toward a principled understanding of underlying computations. Such algorithmic explanations offer a pathway to human-understandable interpretability, enabling comprehension of the model's internal reasoning performance measures. This can in turn lead to more sample-efficient methods for training and improving performance, as well as novel architectures for end-to-end and multi-agent systems.

Related papers

• Fitness Landscape of Large Language Model-Assisted Automated Algorithm Search [15.767411435705752]
  We show and analyze the fitness landscape of Large Language Models-assisted Algorithm Search.<n>Our findings reveal that LAS landscapes are highly multimodal and rugged.<n>We also demonstrate how population size influences exploration-exploitation trade-offs and the evolving trajectory of elite algorithms.
  arXiv  Detail & Related papers  (2025-04-28T09:52:41Z)
• Enhancing LLM Reasoning with Reward-guided Tree Search [95.06503095273395]
  o1-like reasoning approach is challenging, and researchers have been making various attempts to advance this open area of research.<n>We present a preliminary exploration into enhancing the reasoning abilities of LLMs through reward-guided tree search algorithms.
  arXiv  Detail & Related papers  (2024-11-18T16:15:17Z)
• EVOLvE: Evaluating and Optimizing LLMs For Exploration [76.66831821738927]
  Large language models (LLMs) remain under-studied in scenarios requiring optimal decision-making under uncertainty. We measure LLMs' (in)ability to make optimal decisions in bandits, a state-less reinforcement learning setting relevant to many applications. Motivated by the existence of optimal exploration algorithms, we propose efficient ways to integrate this algorithmic knowledge into LLMs.
  arXiv  Detail & Related papers  (2024-10-08T17:54:03Z)
• On the Design and Analysis of LLM-Based Algorithms [74.7126776018275]
  Large language models (LLMs) are used as sub-routines in algorithms. LLMs have achieved remarkable empirical success. Our proposed framework holds promise for advancing LLM-based algorithms.
  arXiv  Detail & Related papers  (2024-07-20T07:39:07Z)
• Large Language Model-Enhanced Algorithm Selection: Towards Comprehensive Algorithm Representation [27.378185644892984]
  This paper introduces Large Language Models (LLMs) into algorithm selection for the first time. LLMs not only captures the structural and semantic aspects of the algorithm, but also demonstrates contextual awareness and library function understanding. The selected algorithm is determined by the matching degree between a given problem and different algorithms.
  arXiv  Detail & Related papers  (2023-11-22T06:23:18Z)
• Neural Algorithmic Reasoning Without Intermediate Supervision [21.852775399735005]
  We focus on learning neural algorithmic reasoning only from the input-output pairs without appealing to the intermediate supervision. We build a self-supervised objective that can regularise intermediate computations of the model without access to the algorithm trajectory. We demonstrate that our approach is competitive to its trajectory-supervised counterpart on tasks from the CLRSic Algorithmic Reasoning Benchmark.
  arXiv  Detail & Related papers  (2023-06-23T09:57:44Z)
• Representation Learning with Multi-Step Inverse Kinematics: An Efficient and Optimal Approach to Rich-Observation RL [106.82295532402335]
  Existing reinforcement learning algorithms suffer from computational intractability, strong statistical assumptions, and suboptimal sample complexity. We provide the first computationally efficient algorithm that attains rate-optimal sample complexity with respect to the desired accuracy level. Our algorithm, MusIK, combines systematic exploration with representation learning based on multi-step inverse kinematics.
  arXiv  Detail & Related papers  (2023-04-12T14:51:47Z)
• Adaptive Discretization in Online Reinforcement Learning [9.560980936110234]
  Two major questions in designing discretization-based algorithms are how to create the discretization and when to refine it. We provide a unified theoretical analysis of tree-based hierarchical partitioning methods for online reinforcement learning. Our algorithms are easily adapted to operating constraints, and our theory provides explicit bounds across each of the three facets.
  arXiv  Detail & Related papers  (2021-10-29T15:06:15Z)
• Identifying Co-Adaptation of Algorithmic and Implementational Innovations in Deep Reinforcement Learning: A Taxonomy and Case Study of Inference-based Algorithms [15.338931971492288]
  We focus on a series of inference-based actor-critic algorithms to decouple their algorithmic innovations and implementation decisions. We identify substantial performance drops whenever implementation details are mismatched for algorithmic choices. Results show which implementation details are co-adapted and co-evolved with algorithms.
  arXiv  Detail & Related papers  (2021-03-31T17:55:20Z)
• Evolving Reinforcement Learning Algorithms [186.62294652057062]
  We propose a method for meta-learning reinforcement learning algorithms. The learned algorithms are domain-agnostic and can generalize to new environments not seen during training. We highlight two learned algorithms which obtain good generalization performance over other classical control tasks, gridworld type tasks, and Atari games.
  arXiv  Detail & Related papers  (2021-01-08T18:55:07Z)
• A Brief Look at Generalization in Visual Meta-Reinforcement Learning [56.50123642237106]
  We evaluate the generalization performance of meta-reinforcement learning algorithms. We find that these algorithms can display strong overfitting when they are evaluated on challenging tasks.
  arXiv  Detail & Related papers  (2020-06-12T15:17:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.