Rhicheek Patra

Richmond Alake, Cesare Bernardis, Paul Cayet et al. · ibm-research

Agent memory is a systems problem for long-horizon agents. Practical deployments require retention of task state across extended conversations, recovery of user-specific facts and preferences across sessions, and accumulation of procedural knowledge from prior outcomes. These requirements extend beyond document retrieval: a memory layer must determine which interactions become durable state, how that state is scoped, how it is retrieved under latency constraints, and how it is revised or removed over time. This report studies Oracle Agent Memory as a database-native memory substrate built on Oracle Database. Three themes organize the discussion: memory as a lifecycle spanning ingestion, extraction, consolidation, retrieval, summarization, and revision or removal; a layered architecture that separates an active memory core from a passive memory-store interface with explicit scope control across users, agents, and threads; and evaluation methodology in which downstream task accuracy is complemented by memory-centric measures such as evidence retrieval, recall, latency, and estimated token use. The report summarizes LongMemEval results, reaching 93.8% accuracy, compares Oracle Agent Memory against flat-history baselines, using about 10.7x fewer tokens, and published or reported external baselines where available, and closes with implementation-oriented appendix material covering setup, thread lifecycle, and search semantics.

Cong Duy Vu Hoang, Gioacchino Tangari, Clemence Lanfranchi et al.

The growing adoption of large language models (LLMs) in business applications has amplified interest in Natural Language to SQL (NL2SQL) solutions, in which there is competing demand for high performance and efficiency. Domain- and customer-specific requirements further complicate the problem. To address this conundrum, we introduce Distill-C, a distilled customization framework tailored for NL2SQL tasks. Distill-C utilizes large teacher LLMs to produce high-quality synthetic data through a robust and scalable pipeline. Finetuning smaller and open-source LLMs on this synthesized data enables them to rival or outperform teacher models an order of magnitude larger. Evaluated on multiple challenging benchmarks, Distill-C achieves an average improvement of 36% in execution accuracy compared to the base models from three distinct LLM families. Additionally, on three internal customer benchmarks, Distill-C demonstrates a 22.6% performance improvement over the base models. Our results demonstrate that Distill-C is an effective, high-performing and generalizable approach for deploying lightweight yet powerful NL2SQL models, delivering exceptional accuracies while maintaining low computational cost.

Soufiane Amini, Yassine Benajiba, Cesare Bernardis et al. · ibm-research

The proliferation of agent frameworks has led to fragmentation in how agents are defined, executed, and evaluated. Existing systems differ in their abstractions, data flow semantics, and tool integrations, making it difficult to share or reproduce workflows. We introduce Open Agent Specification (Agent Spec), a declarative language that defines AI agents and agentic workflows in a way that is compatible across frameworks, promoting reusability, portability and interoperability of AI agents. Agent Spec defines a common set of components, control and data flow semantics, and schemas that allow an agent to be defined once and executed across different runtimes. Agent Spec also introduces a standardized Evaluation harness to assess agent behavior and agentic workflows across runtimes - analogous to how HELM and related harnesses standardized LLM evaluation - so that performance, robustness, and efficiency can be compared consistently across frameworks. We demonstrate this using four distinct runtimes (LangGraph, CrewAI, AutoGen, and WayFlow) evaluated over three different benchmarks (SimpleQA Verified, $τ^2$-Bench and BIRD-SQL). We provide accompanying toolsets: a Python SDK (PyAgentSpec), a reference runtime (WayFlow), and adapters for popular frameworks (e.g., LangGraph, AutoGen, CrewAI). Agent Spec bridges the gap between model-centric and agent-centric standardization & evaluation, laying the groundwork for reliable, reusable, and portable agentic systems.

Georgios Damaskinos, El Mahdi El Mhamdi, Rachid Guerraoui et al.

Asynchronous distributed machine learning solutions have proven very effective so far, but always assuming perfectly functioning workers. In practice, some of the workers can however exhibit Byzantine behavior, caused by hardware failures, software bugs, corrupt data, or even malicious attacks. We introduce \emph{Kardam}, the first distributed asynchronous stochastic gradient descent (SGD) algorithm that copes with Byzantine workers. Kardam consists of two complementary components: a filtering and a dampening component. The first is scalar-based and ensures resilience against $\frac{1}{3}$ Byzantine workers. Essentially, this filter leverages the Lipschitzness of cost functions and acts as a self-stabilizer against Byzantine workers that would attempt to corrupt the progress of SGD. The dampening component bounds the convergence rate by adjusting to stale information through a generic gradient weighting scheme. We prove that Kardam guarantees almost sure convergence in the presence of asynchrony and Byzantine behavior, and we derive its convergence rate. We evaluate Kardam on the CIFAR-100 and EMNIST datasets and measure its overhead with respect to non Byzantine-resilient solutions. We empirically show that Kardam does not introduce additional noise to the learning procedure but does induce a slowdown (the cost of Byzantine resilience) that we both theoretically and empirically show to be less than $f/n$, where $f$ is the number of Byzantine failures tolerated and $n$ the total number of workers. Interestingly, we also empirically observe that the dampening component is interesting in its own right for it enables to build an SGD algorithm that outperforms alternative staleness-aware asynchronous competitors in environments with honest workers.

Rachid Guerraoui, Erwan Le Merrer, Rhicheek Patra et al.

Recommenders personalize the web content by typically using collaborative filtering to relate users (or items) based on explicit feedback, e.g., ratings. The difficulty of collecting this feedback has recently motivated to consider implicit feedback (e.g., item consumption along with the corresponding time). In this paper, we introduce the notion of consumed item pack (CIP) which enables to link users (or items) based on their implicit analogous consumption behavior. Our proposal is generic, and we show that it captures three novel implicit recommenders: a user-based (CIP-U), an item-based (CIP-I), and a word embedding-based (DEEPCIP), as well as a state-of-the-art technique using implicit feedback (FISM). We show that our recommenders handle incremental updates incorporating freshly consumed items. We demonstrate that all three recommenders provide a recommendation quality that is competitive with state-of-the-art ones, including one incorporating both explicit and implicit feedback.

Rhicheek Patra, Egor Samosvat, Michael Roizner et al.

Recommenders have become widely popular in recent years because of their broader applicability in many e-commerce applications. These applications rely on recommenders for generating advertisements for various offers or providing content recommendations. However, the quality of the generated recommendations depends on user features (like demography, temporality), offer features (like popularity, price), and user-offer features (like implicit or explicit feedback). Current state-of-the-art recommenders do not explore such diverse features concurrently while generating the recommendations. In this paper, we first introduce the notion of Trackers which enables us to capture the above-mentioned features and thus incorporate users' online behaviour through statistical aggregates of different features (demography, temporality, popularity, price). We also show how to capture offer-to-offer relations, based on their consumption sequence, leveraging neural embeddings for offers in our Offer2Vec algorithm. We then introduce BoostJet, a novel recommender which integrates the Trackers along with the neural embeddings using MatrixNet, an efficient distributed implementation of gradient boosted decision tree, to improve the recommendation quality significantly. We provide an in-depth evaluation of BoostJet on Yandex's dataset, collecting online behaviour from tens of millions of online users, to demonstrate the practicality of BoostJet in terms of recommendation quality as well as scalability.