arXiv subjects

Sarthak Ghosh

Publications and source records attributed to Sarthak Ghosh.

4 recordsLinked to original sources

arXiv2026

SLATE: A Large-Scale Benchmark for Long-Horizon API Selection in E-commerce Tool Use

Enterprises increasingly want LLM agents to automate standard operating procedures over their software stacks -- mapping each step of a procedure to the right call among thousands of internal APIs, without hand-authored step-to-API mappings. The core difficulty is API selection at scale: choosing and composing the correct tools over a long horizon, judged only by whether the overall task succeeds. Current tool-use benchmarks do not test this regime -- they expose only a handful of tools, evaluate single-step invocation, or grade multi-step behavior with subjective LLM-as-a-judge scores that penalize valid alternative trajectories. We introduce SLATE (Synthetic Large-scale API Toolkit for E-commerce), a large-scale benchmark that tests large-toolset, long-horizon API selection in a realistic enterprise domain. SLATE couples (i) queries decomposed into hierarchical, tool-grounded procedures over a ~1,000-tool e-commerce library, (ii) a deterministic, context-aware simulator that returns recorded outputs for matching calls and default responses otherwise, and (iii) an end-to-end task-success metric -- a task succeeds when it reaches every reference outcome -- that credits diverse-but-valid trajectories while remaining fully automatic and reproducible. Experiments with six open decision models and the language models they are built on, from 3B to 12B parameters, show that even the best model completes only 5.0% of the tasks, establishing SLATE as a challenging testbed for reliable enterprise tool use. We release SLATE, its simulator, and the evaluation harness at https://geshijoker.github.io/slate/

cs.CL↗

arXiv2026

WASP-Mem: Let Agent Organize Procedural Memory. Human Spend Effort on Diagnosis

LLM agents struggle to automate complex enterprise procedures because static, pre-built harness cannot capture proprietary policies, undocumented behaviors, or state-dependent recovery paths that surface only during execution. Procedural memory accrued from agent experience offers a training-free path to closing this gap -- yet unlike episodic records or semantic facts, it must consolidate repeated successes and diagnosed failures into reusable actions and recovery strategies. We introduce WASP-Mem (Wiki-based Agent-Synthesized Procedural Memory), a nimble but effective method that distills execution traces and feedback into persistent, interlinked wiki pages, enabling continual agent learning. On TheAgentCompany benchmark, WASP-Mem raises the checkpoint pass rate by 4.6-6.1% over the strongest baseline we compared across three models while requiring 57-61% less storage. We further present a systematic controlled study spanning consolidation method, feedback type, trace ingestion scope, and refinement rounds. Our results reveal that prescriptive knowledge organization such as human defined schemas and explicit deduplication rules hurts performance rather than closing the procedural-knowledge gap. In contrast, human diagnosis converts failed traces into corrective knowledge and drives sustained improvement: a single round of human diagnosis outperforms four rounds of self-judgment by 7.0-10.9%. These findings point to a simple division of labor: let agents organize procedural memory, and invest human effort where it matters most -- diagnosis.

cs.AI↗

arXiv2026

Long-Horizon Plan Execution in Large Tool Spaces through Entropy-Guided Branching

Large Language Models (LLMs) have significantly advanced tool-augmented agents, enabling autonomous reasoning via API interactions. However, executing multi-step tasks within massive tool libraries remains challenging due to two critical bottlenecks: (1) the absence of rigorous, plan-level evaluation frameworks and (2) the computational demand of exploring vast decision spaces stemming from large toolsets and long-horizon planning. To bridge these gaps, we first introduce SLATE (Synthetic Large-scale API Toolkit for E-commerce), a large-scale context-aware benchmark designed for the automated assessment of tool-integrated agents. Unlike static metrics, SLATE accommodates diverse yet functionally valid execution trajectories, revealing that current agents struggle with self-correction and search efficiency. Motivated by these findings, we next propose Entropy-Guided Branching (EGB), an uncertainty-aware search algorithm that dynamically expands decision branches where predictive entropy is high. EGB optimizes the exploration-exploitation trade-off, significantly enhancing both task success rates and computational efficiency. Extensive experiments on SLATE demonstrate that our dual contribution provides a robust foundation for developing reliable and scalable LLM agents in tool-rich environments.

cs.AI↗

arXiv2019

Value of Information in Probabilistic Logic Programs

In medical decision making, we have to choose among several expensive diagnostic tests such that the certainty about a patient's health is maximized while remaining within the bounds of resources like time and money. The expected increase in certainty in the patient's condition due to performing a test is called the value of information (VoI) for that test. In general, VoI relates to acquiring additional information to improve decision-making based on probabilistic reasoning in an uncertain system. This paper presents a framework for acquiring information based on VoI in uncertain systems modeled as Probabilistic Logic Programs (PLPs). Optimal decision-making in uncertain systems modeled as PLPs have already been studied before. But, acquiring additional information to further improve the results of making the optimal decision has remained open in this context. We model decision-making in an uncertain system with a PLP and a set of top-level queries, with a set of utility measures over the distributions of these queries. The PLP is annotated with a set of atoms labeled as "observable"; in the medical diagnosis example, the observable atoms will be results of diagnostic tests. Each observable atom has an associated cost. This setting of optimally selecting observations based on VoI is more general than that considered by any prior work. Given a limited budget, optimally choosing observable atoms based on VoI is intractable in general. We give a greedy algorithm for constructing a "conditional plan" of observations: a schedule where the selection of what atom to observe next depends on earlier observations. We show that, preempting the algorithm anytime before completion provides a usable result, the result improves over time, and, in the absence of a well-defined budget, converges to the optimal solution.

cs.AI↗