small language model

Guarantees, dispositions and answerable judges: An empirical basis for the behavioural certification of tool-using language agents

Buyers of tool-using language agents ask for evidence that an agent respects the authority it has been given, and what they receive is usually the vendor's own account. We ask what an independent behavioural certificate can claim about an agent the certifier cannot inspect, using a deployed certification suite as the i...

Applying the DFA-WK method to evaluate long-context retention in large language models

The growth of context windows in Large Language Models (LLMs), now exceeding millions of tokens, has made evaluating long-range dependency retention a central challenge. Traditional evaluation approaches, such as pointwise retrieval tests (needle-in-a-haystack) and average perplexity, capture only part of this dynamic....

The Effectiveness of Early Intervention Programs for Children Aged 0–3: A Systematic Review and Meta-Analysis

Background: Existing syntheses on infant early interventions focus on single-risk cohorts and overlook implementation moderators. To address these gaps, this systematic review and meta-analysis adopts a transdiagnostic approach to evaluate interventions across diverse risk profiles and identify predictive intervention...

On the Feasibility of Distilled Small Language Models for Agentic Text-to-SQL in Portuguese Under Real-World Conditions

Ankyra: Sound Neuro-Symbolic Reasoning with Verifiable Proofs

We present Ankyra, a domain-general neuro-symbolic reasoning system built on a strict separation of roles: the language model \emph{proposes} a formalization, and a deterministic symbolic engine \emph{decides} it. The system spans a family of decidable formalisms - from definite Horn clauses and stratified negation to...

MolecureAI: an agentic AI platform for autonomous drug discovery

Accelerating the identification of viable drug candidates across therapeutic areas demands autonomous systems for molecular generation that enforce chemical correctness by structural invariants rather than model capability. We present MolecureAI, an autonomous AI agent architecture for closed-loop drug discovery with c...

Siesta: Identitas yang Koheren pada Model Bahasa Orchestrator dengan Kepribadian Emergen — Human-Parity Orchestration (HuPO)

Siesta is a small-scale independent research project exploring identity coherence in an Indonesian-language orchestrator model — a "brain" that stays behaviorally consistent while delegating specialized tasks to separate hidden models, following an architectural pattern we call Human-Parity Orchestration (HuPO). Built...

Pengembangan Lembar Kerja Peserta Didik (LKPD) Berbasis Kearifan Lokal Pada Pembelajaran IPAS Kelas V SD Negeri 33 Lubuk Linggau

This study aims to develop a local wisdom-based Student Worksheet (LKPD) for Grade V IPAS learning at SD Negeri 33 Lubuk Linggau and to determine its validity, practicality, and effectiveness. This study employed a Research and Development (R&D) method using the ADDIE model, consisting of Analysis, Design, Development,...

Efficiency, Cost, and Quality: Estimating Token Consumption as a Process Metric for AI-Native Software Projects

Efficiency, Cost, and Quality: Estimating Token Consumption as a Process Metric for AI-Native Software Projects Author: Wenchao Shi Date: September 2026 Keywords: AI-native development, token accounting, process metrics, cost of quality, statistical process control, ISO 9001, COCOMO, evidence-based management Abstract...

Prompting Strategy Evaluation for Offline Code Generation Agents on Resource-Constrained Hardware: A Comparative Study of Sub-2B Parameter Models

The deployment of large language model (LLM)-based code generation agents typicallyrelies on cloud infrastructure, limiting their applicability in air-gapped, low-resource, oroffline environments. This paper investigates the feasibility and effectiveness of runningsub-2B parameter LLMs as offline code generation agents...

Easier Said Than Done: Strangers, Impostors, and the Negatives That Teach a Small Learner to Say "I Don't Know"

How do you teach a model to say "I don't know"? The usual recipe collects questions it cannot answer, trains it to refuse them, and checks on a held-out part of the same collection, which cannot tell a learned rule from a memorized list, nor say which kind of unknown the rule covers. A question can carry a name the mod...

Prompting Strategy Evaluation for Offline Code Generation Agents on Resource-Constrained Hardware: A Comparative Study of Sub-2B Parameter Models

The deployment of large language model (LLM)-based code generation agents typicallyrelies on cloud infrastructure, limiting their applicability in air-gapped, low-resource, oroffline environments. This paper investigates the feasibility and effectiveness of runningsub-2B parameter LLMs as offline code generation agents...