Small Language Models

The scaling laws hold that a language model grows more capable with more parameters and more training data, and Mixture-of-Experts (MoE) architectures have ridden these laws to remarkable results, activating only a fraction of an enormous stored parameter bank for each token. That success is built on static pretraining data. A deployed model faces a different world, where much of the data that would make it more useful is not in its training set but in the live interaction it is currently handling, such as the facts a user supplies or the corrections they give. A conventional model cannot learn from this data, because its weights are frozen after training. Instead, the knowledge and behaviour supplied at run time are placed in the prompt, by retrieval or instruction, and re-read on every request only to be discarded once the request ends. We ask how an architecture could learn from live interaction by writing it into its weights. Taking inspiration from MoE, we propose the \textbf{Infinite-Parameter LLM}. A compact hypernetwork turns the data given at run time into a low-rank modulation of a shared base network, so the feed-forward weights are generated from live data rather than stored in a fixed bank. Where prior weight generators read the context once and freeze, we carry a Bayesian belief over the generator's latent code and update it online, so the effective weight is re-derived from that evolving belief as the session proceeds rather than fixed after one read. The stored footprint stays fixed, yet the weights the model can compile are effectively infinite. For the knowledge and behaviour supplied at run time, carrying them in the weights rather than the prompt is amortized in compute, frees the context window, persists across turns, and can generalise better than in-context use. We specify an evaluation protocol that tests exactly this against in-context learning and retrieval.

Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model's induction heads.

Dense retrieval systems increasingly rely on multi-billion-parameter language models, whose memory and computational requirements make large-scale indexing, frequent corpus updates, and low-latency serving costly. We present a three-stage training pipeline for developing compact and efficient retrievers that remain competitive with substantially larger models. The pipeline combines cross-lingual alignment, relational knowledge distillation, and contrastive fine-tuning. It requires no original ground-truth relevance labels, relying exclusively on supervision generated by strong embedding models and rerankers utilised as teachers. Using this pipeline, we develop PolDense and EuroDense, both supporting contexts of up to 8,192 tokens. PolDense is a family of six Polish retrievers ranging from 17M to 1B parameters. EuroDense is a 435M-parameter retriever supporting nine European languages. We conduct an extensive evaluation covering 41 Polish and 150 multilingual retrieval tasks. The results demonstrate strong quality-efficiency trade-offs. PolDense-1B outperforms the evaluated retrievers with up to 9B parameters, while the PolDense family forms the Pareto frontier across model sizes. Among the evaluated models below 1B parameters, EuroDense ranks first in both task-averaged and language-averaged performance and leads in seven of nine languages. We release all models publicly.

The proliferation of highly capable open-weight Small Language Models (SLMs) democratizes access to advanced cybersecurity capabilities, posing an escalating risk as these models can bypass proprietary API guardrails when deployed locally. However, SLMs deployed as autonomous agents often struggle with long-horizon, exploratory tasks like cybersecurity Capture The Flag (CTF) challenges due to context bloat and cognitive degradation from accumulated tool-call outputs. To understand and mitigate this cybersecurity threat, we introduce context segmentation, a two-level agentic framework that divides complex exploitation tasks into manageable, contextually isolated sub-problems. Evaluating on the picoCTF dataset using memory-constrained gemma-4 models, we demonstrate that for the E4B model, our strategy acts as an intelligent search, achieving competitive rewards with superior token efficiency compared to brute-force retries, and successfully solving 18.52% of tasks that standard agentic execution fails to complete. Code is available at https://github.com/9xeb/context-segmentation.

A language model normally begins training with random word embeddings: whatever 'banana' means must be learned from training corpora. I implement St. Augustine's picture of word learning, meaning by ostension, for a small masked language model (DeBERTa) trained on 10M words: before training, visually grounded tokens receive embeddings derived from the image regions they label; other tokens start random. Visual initialization leaves a measurable imprint that lasts until the end of training. At the same time, the effect remains invisible under most BabyLM benchmarks, which probe abstract grammatical knowledge: visual initialization does not affect performance there. The only zero-shot exception is object-property knowledge (COMPS, Misra et al. 2023), where seeding helps in every configuration. To follow up on this result, I build a corpus-tailored version of the Visual-Property Swap benchmark (Lin et al., 2026), which tests color, material, size, and shape knowledge, with per-item training frequency and seeded status. Here, vision-seeded models have a persistent, seed- replicated advantage, confined to the seeded words. As a causal test, I show that synthetic grounding of previously unseeded words transfers the advantage to exactly those words. Function words and abstract vocabulary also receive strong visual seeds and retain them throughout training, and the training objective draws on them: held-out mask-prediction loss falls for these words in every seed. However, no benchmark I run registers this. What evaluation would pick this up remains an open question.

Ultra-low-bit language models promise reductions in storage and memory traffic, but a nominal "1.58-bit" label does not specify the deployed representation or its execution cost. We study a scale-up of an aggressive post-training conversion pipeline from Qwen3-4B to Qwen3-8B. The conversion uses KOTMS rotation, E2M-ATQ adaptive ternarisation, and GPTQ-style error compensation in a weight-only A16 configuration. We do not claim these algorithms as new. Our contribution is the end-to-end scale-up characterisation: an external reproduction gate, matched 4B/8B capability analysis, cross-corpus perplexity, effective-bit accounting, lossless lattice-aware packing, and direct packed execution. The 8B model reaches a three-corpus perplexity ratio of 1.361x, with WikiText-2, C4, and PTB ratios of 1.318x, 1.393x, and 1.371x. On eight zero-shot tasks at n = 500, mean accuracy is 64.6% versus 72.4% for FP16, corresponding to 78.5% chance-corrected retention and a 7.8-point absolute cost. The matched 4B run retains 69.6%, yielding an 8.9-point 8B advantage. The packed checkpoint is 8.24 GiB and preserves the recorded perplexity to measurement precision. Direct packed execution reaches 15.52 tokens/s in 7.35 GiB, while a preliminary packed GEMV remains slower than FP16 cuBLAS. The result is a validated scale-up baseline: model size improves robustness to aggressive post-training discretisation, actual serialisation is solved for the measured artefact, and direct execution is feasible, while broader seeds, calibration distributions, and kernel optimisation remain open.

We study how physical-state inputs affect a 0.8B hybrid language model adapted for manipulation with 6.2M trainable parameters. Six conditions are trained on three LIBERO-Spatial tasks and evaluated over three seeds and 540 held-out rollouts. Conditioning recurrent decay gates on geometric increments yields 28.9% success, compared with 36.7% when those increments are shuffled during training and 24.4% without explicit object/goal geometry. Both geometry policies receive correct inputs at evaluation. A token adapter using the same increments scores 27.8%; differences vary across seeds and remain inconclusive. Token-clock conditioning scores 11.1%, including one seed that fails to converge. In separate robustness tests, a state-only relative-coordinate policy retains 7/10 success under frame relabeling, whereas all four tested visual policies fall to at most 3/20 after a 5 cm object displacement. These results show no reliable advantage from training-time geometric alignment under this recipe and illustrate the gap between coordinate invariance and physical-layout generalization. Episode records, seed-level analyses, and figure-generation code accompany the paper.

In-vehicle assistants must translate natural-language requests into accurate vehicle function calls under strict memory and latency constraints, making small language models (SLMs) attractive for on-device deployment. For such models, a key design choice is how the available function surface is presented. Two approaches are to represent each function with a dedicated Functional Token (FT) or provide function schemas directly in the prompt. FTs enable compact inference but are restricted to functions learned during training, whereas Schema-in-Prompt (SIP) can generalize to unseen functions at the cost of longer prompts and higher inference overhead. We introduce a benchmark of 9,822 single-turn examples spanning 79 vehicle functions derived from Android Automotive, including held-out functions and requests requiring refusal. We compare both approaches under matched fine-tuning across four SLMs from 270M to 1.7B parameters. On functions seen during training, scaling provides limited benefit: the 270M model can match the 1.7B model, while the strongest overall performance occurs at 0.6B. On held-out functions, FT achieves zero accuracy by construction, whereas SIP generalizes and improves substantially with scale. On out-of-scope requests, FT can invoke an unavailable function it was trained to emit, while SIP more reliably refuses based on the functions offered. This flexibility comes with higher memory use and latency. Our theoretical analysis explains how SIP enables generalization and why longer schema contexts increase inference cost. Overall, function-surface representation, rather than model scale alone, determines the capabilities and failure modes of SLM-based vehicle function calling.

We approach a cognitive simulation perspective on episodic and semantic memory in multiple-choice question answering by incorporating text from individual text corpora (ITC) into retrieval-augmented generation and DoRA fine-tuning. We web-crawl the search histories of 515 participants who answered 36 multiple-choice knowledge items and analyze a stratified subsample of 150 participants. For each participant, one DoRA adapter consolidates their ITC into a small language model (SLM) whose baseline correctness falls below the participants' lowest quartile. The adapter measurably writes the ITC into the weights: it fits its own participant's held-out text better than other participants' texts (dz =1.27), an individuality effect that increases with ITC size in rank order. On the generalized knowledge test, however, the adapter adds knowledge rather than alignment with the individual: log-loss match improves, whereas match accuracy under a bias-corrected PMI readout does not, and retrieval adds nothing on top. Our results demonstrate that ITCs can be consolidated into the weights of SLMs, an encouraging basis for individualized tutoring agents, and we discuss how to move from there toward a realistic simulation of episodic and semantic memory at the individual level.

Resource constrained single-board computers including Raspberry Pi, NVIDIA Jetson Nano, Arduino UNO Q, Orange Pi, and LattePanda motivate on-device small language model (SLM) agents that reduce cloud dependence, improve data locality, and tolerate intermittent connectivity. Model Context Protocol (MCP)-style tool invocation demands more than fluent generation: an agent must emit machine-readable JSON, select the correct tool, supply all required arguments, and avoid unintended actions. We establish a platform-agnostic CPU baseline by evaluating five open-weight models below two billion parameters Phi-1.5, Pythia-1.4B, TinyLlama-1.1B-Chat, Qwen2.5-0.5B, and Qwen2.5-1.5B on 100 prompts spanning weather retrieval, web search, calculation, email composition, and task creation, under greedy decoding and nucleus sampling. A recovery parser strips Markdown fences, extracts brace-delimited substrings, and scores parseability, tool-name correctness, argument completeness, and value agreement. Under this criterion, Qwen2.5-1.5B achieves 75% (greedy) and 79% (sampling); Qwen2.5-0.5B achieves 72% (greedy) but drops to 32% under sampling. Phi-1.5 scores 0%; Pythia and TinyLlama reach at most 7%. A strict post-hoc audit finds only 5 of 1,000 raw responses directly parseable as JSON, exposing near-total dependence on output recovery. A CPU resource probe shows Qwen2.5-1.5B requires 7,960 MiB and 30.782 s mean latency; Qwen2.5-0.5B uses 3,637 MiB and 10.627 s, revealing a reliability-resource trade-off for edge deployment. These results do not cover the named boards directly or a full MCP implementation. Safe deployment requires schema validation, constrained generation, least-privilege execution, and human escalation for consequential actions.

Ultra-low-bit language models can reduce storage and memory bandwidth, but a nominal "1.58-bit" label does not fully describe the stored representation, retained capability, or runtime behavior. We study an end-to-end post-training conversion of Qwen, an instruction-tuned 4B-parameter model, using KOTMS rotation, E2M-ATQ ternarization, and GPTQ-style error compensation from TWLA. The experiment is weight-only: activations remain at 16-bit precision, so ILA-AMP is omitted. We evaluate effective bit accounting, task capability retention, perplexity, calibration sensitivity, checkpoint composition, and deployment behavior. The final conversion uses 1.641 effective bits per weight for quantized linear weights, with 81.62% of model parameters targeted. Across ten scored capability comparisons, accuracy falls from 64.5% to 54.7%. Degradation is uneven: BoolQ retains 84.6% chance-corrected teacher performance, while ARC-Challenge retains 43.8%. Perplexity rises from 13.639 to 18.748 on WikiText-2, 24.700 to 31.992 on PTB, and 19.831 to 28.966 on C4. A subsequent packing run preserves the ternary planes and scales, reducing reported model size from 8.29 GiB to 3.96 GiB with essentially unchanged perplexity. A separate third-party packing attempt was lossy and is excluded from the primary artifact claim. The packed artifact has not been benchmarked end-to-end for task accuracy or generation throughput. A preliminary Triton GEMV microbenchmark is 4.6x slower than FP16 cuBLAS on one tested shape. We therefore do not claim that compression alone yields faster inference.

Edge-AI model selection is commonly driven by one isolated metric - accuracy, latency, memory, energy, or safety, even though a deployable language model must balance all five. Our work focuses on answering the question whether na- tively trained small language models (SLMs) or large language models (LLMs) compressed through post-training quantization offer the more sustainable edge- deployment trade-off. We introduce a reproducible Holistic Sustainability Score (HSS) organized around the triple bottom line: an economic pillar for capability and systems efficiency, an environmental pillar for operational GPU energy and a social pillar for harmful-prompt robustness. Five BF16 SLMs and five LLMs under different quantization approaches - BF16, INT8, NF4 4-bit, GPTQ 4-bit, and GGUF Q4 produce 30 measured configurations. Capability is assessed on five zero-shot benchmarks; efficiency uses latency, throughput, peak VRAM and energy; and safety is approximated by attack success rate on five harmful prompts. Qwen3-30B-A3B/GGUF Q4 ranks first in the combined pool (93.38), followed by Mistral-Small-24B/GGUF Q4 (92.40), while Phi-4-mini/BF16 is the highest- ranked SLM in that pool (89.49). Thus, the hypothesis that native SLMs must be the most sustainable edge choice is not supported universally; optimized quantized LLMs can win overall, while SLMs remain competitive through lower resource demand. Quantization is a systems-level choice rather than a monotonic precision- efficiency trade-off and HSS remains relative to its comparison pool and proxy definitions.

For emerging scientific research domains, local Small Language Models (SLMs) are becoming more attractive, as they offer stronger privacy control and more stable deployment pipelines than Large Language Models. However, in practice, scientific question-answering on SLMs often operates under inevitable constraints: small literature collections, fragmented evidence, limited context window and reasoning abilities. We propose the Evidence-Grounded Typed Knowledge Graph (EGT-KG), a retrieval framework to improve information retrieval with local SLMs. We assessed three question-answering settings: a vanilla Retrieval-Augmented Generation (RAG) workflow and two EGT-KG workflows: an automatically generated relation schema (AS) and an expert-defined relation schema (ES). Our experiments were evaluated with a six-dimensional evaluation framework (S3CRF: Soundness, Correctness, Completeness, Conciseness, Relevance, Fluency) on a Biopolymer-bound Soil Composite literature benchmark, showing that EGT-KG outperforms the vanilla RAG method in most settings, with the best improvement from llama3:8b: a Final Score of 70.37 (+14.67%) and 68.82 (+12.14%) by AS/ES EGT-KG variants.

Assessing claim check-worthiness is an essential first step in automated fact-checking pipelines. This work is motivated by a real deployment challenge at an early-stage startup: running large language models (LLMs) over every incoming claim is cost- and latency-prohibitive, yet smaller models sacrifice accuracy. We propose NN-PPI, a pointwise extension of Prediction-Powered Inference (PPI) that calibrates model predictions at inference time as a lightweight post-hoc layer, without re-training the underlying model. NN-PPI achieves weighted F1 gains ranging from 12% to 33.80% depending on the size and performance of the baseline model, bringing SLMs on par with larger LLMs. Beyond few-shot SLMs, NN-PPI further improves a production-deployed fine-tuned model, demonstrating that residual calibration is complementary to supervised fine-tuning. By recovering LLM-level accuracy from models that are an order of magnitude cheaper to serve, it makes accurate check-worthiness detection substantially cheaper to operate at scale. Our code and data can be found at https://anonymous.4open.science/r/arr-claim-worthiness-F237.

Rubric-based reinforcement learning extends RL beyond tasks with exact answers or rule-based verifiers by scoring responses against instance-specific criteria. However, this makes reward computation expensive: training requires repeated rubric judging, often with proprietary APIs or local generative LLM judges with 7B parameters or more. We study whether smaller language models can serve as efficient and reliable rubric-based judges. To make this question measurable, we construct PointRubric and RaR-Science-Static, two pointwise rubric-based evaluation datasets with instance-specific criteria and itemwise satisfaction labels. We compare three ways of extracting criterion-level judgments from small models: Generative verdicts, Yes/No Logprob margins, and Probe judges. Across both datasets, the Qwen3-1.7B Probe judge achieves the strongest criterion-level agreement among these methods, outperforming Generative and Logprob judges. Used as a GRPO reward model, it trains a policy from 0.232 to 0.643 on RaR-Science rubric score, compared with 0.594 for an 8B Generative judge baseline, while the baseline requires 10.7$\times$ more reward-judge time. Task and domain transfer experiments further suggest that Probe judges preserve criterion-level reward structure across settings.

We show that sequence-level distillation from a capable long-context teacher model is a simple, annotation-free, and data-efficient strategy for improving argument saliency coverage in long legal opinion summarization, where small LLMs often struggle to retain the most salient argumentative content. Across student model sizes, distillation consistently surpasses tuning on expert-written summaries in our legal-opinion setting. We further demonstrate that most gains are achieved with as few as ~10 training summaries, highlighting the strong data efficiency of teacher-generated supervision. Finally, we find that summary distillation is sufficient for improvements: reasoning-chain distillation remains competitive with summary-only distillation, but provides marginal benefit when combined with summary supervision.

Embodied intelligent virtual agents are expected to operate as persistent, adaptive, and context-aware entities within complex virtual and Metaverse worlds. However, implementing cognitively capable agents in such environments is conceptually and technologically challenging. Among a range of blueprints and development approaches, the Cognitive Embodied Agent Architecture (CEAA) has been developed as an implementation-oriented framework for architecting components of perception, memory, reasoning, planning, and embodied action. Considering the recent advances in edge computing and generative AI language models, this paper explores the use of Small Language Models (SLMs) to support edge-based operation of selected CEAA components, focusing on "Think" and "Memory" as processes central to cognitive orchestration and persistence of virtual agents in interactive virtual worlds. An edge-based virtual agent gateway system was developed and evaluated on an NVIDIA Jetson Orin NX using Qwen2.5 models of different sizes, exploring the system's capability to process service requests and handle memory-driven conversations. A series of simulation experiments evaluated routing accuracy, memory-read performance, and latency, demonstrating an SLM-driven prototype agent system that partially implements selected CEAA processes to support the development of embodied agents whose cognitive "brain" can operate efficiently and contextually for interactive experiences in immersive virtual worlds.

Small Language Models (SLMs) have emerged as a more efficient alternative to traditional Large Language Models (LLMs), offering promising potential in resource-constrained scenarios. Existing approaches to building SLMs typically follow two paths: training compact models from scratch, or compressing larger pre-trained models using methods such as pruning, quantization, or distillation. As language models become increasingly integrated into real-world applications, ensuring their trustworthiness has become a critical concern. However, how to build trustworthy SLMs remains an underexplored question. In this work, we present a comprehensive evaluation of SLM trustworthiness across multiple dimensions, including fairness, robustness, privacy, and ethics. We first examine the effects of pruning and quantization, and find that quantization is significantly more effective in preserving trustworthiness compared to pruning. More importantly, we demonstrate that compressing a reliable large model via quantization can produce SLMs with superior trustworthiness and adaptability compared to using small models trained from scratch. Furthermore, knowledge distillation from trustworthy teacher models can further enhance the reliability of SLMs. We hope our findings provide practical guidance and a foundation for future research into the development and deployment of trustworthy small language models.

Multilingual short-text classification supports operational systems such as content moderation, customer support routing, and intent recognition, yet aggregate evaluation often hides large differences between high-resource and low-resource languages. Uniform inference policies are simple to deploy, but they assume that all languages are equally well served. In this work, we evaluate a fixed-list routing strategy that keeps stronger languages on a direct multilingual path and selectively sends weaker languages through translation into English before zero-shot classification. The pipeline is fully self-hosted, uses pretrained compact sentence encoders, and requires no task-specific fine-tuning. We test the approach on two benchmarks chosen to differ in scale and label granularity: a 15-language subset of SIB-200 for seven-way topic classification and a 15-locale subset of MASSIVE for intent classification over an official 60-intent inventory. On SIB-200, the best overall configuration is R1, which translates only the low-resource tier: high-tier and mid-tier Macro-F1 remain unchanged, while low-tier Macro-F1 rises from 0.4632 to 0.6828. On the MASSIVE subset, the same low-tier intervention raises low-tier Macro-F1 from 0.2143 to 0.4417, but the best overall result is obtained by full translation, R3, at Macro-F1 0.4647. Across these two benchmarks, selective translation is a reliable intervention for weaker languages, whereas the optimal routing boundary depends on the task. We therefore report routing through tier-level quality gains and tier-level latency rather than a single global efficiency score.

Agent skills represent a standardized format for packaging procedural knowledge and domain expertise, serving within agent harness systems as an essential mechanism to continually constrain a language model's behavior space for repeatable, high-quality task execution. However, because strong closed-source models entail high inference costs, current popular agent harnesses, such as Codex and OpenClaw, remain prohibitively expensive when deploying these skills to accomplish real-world tasks. The rapid capability enhancement of open-source models deployable on consumer-grade GPUs presents a compelling opportunity to drastically reduce these costs by leveraging skill-based behavioral constraints. Nevertheless, automatically generating effective skills tailored specifically for such compact models remains a significant practical challenge. To address this, we propose SKILLER, a natural-language-driven reinforcement learning framework designed to automatically generate executor-specific skills for small models, which employs a strong model as the actor and critic, treats the small-model agent system as the environment, and propagates all reinforcement learning signals entirely via natural language. Extensive experimental evaluations across five relevant benchmarks using Qwen3.5-9B and Qwen3.5-4B demonstrate that SKILLER outperforms three open-source and one closed-source skill generation or evolution methods, achieving absolute gains ranging from 4.3 to 20.4 percentage points for the 9B model and 1.8 to 13.3 points for the 4B model, while remarkably matching the performance of strong closed-source models on single-skill tasks in SkillsBench. The project is available at https://github.com/DANG-ai/SKILLER.

Deploying large language models (LLMs) for decision support in emergency departments (EDs) faces two major challenges: privacy risks of transmitting patient data to closed-source commercial LLMs and the lack of systematic evaluation of fine-tuning strategies for locally deployable open-source small language models (SLMs). We benchmarked eight open-source SLMs using zero-shot prompting, prefix tuning, Low-Rank Adaptation (LoRA), and full fine-tuning on three ED tasks: triage level prediction, specialist referral recommendation, and diagnosis prediction. Using 2,083 MIMIC-IV-ED cases and Claude Haiku 4.5 and Claude Sonnet 4.5 as baselines, we found that LoRA fine-tuned open-source SLMs outperform commercial baselines on triage level prediction and specialist referral recommendation, while diagnosis prediction remains challenging for open-source SLMs. Confusion matrix analysis further shows that fine-tuned open-source SLMs can detect highest-severity patients missed by the commercial baselines. These results demonstrate that locally deployable SLMs can achieve clinically competitive performance for ED decision support.

Prospective daily symptom tracking is central to premenstrual health assessment, but repeated ordinal forms impose substantial response burden. We formulate conversational administration as an ordinal label-recovery problem: the system actively elicits a small set of symptom clusters and maps each response to the original severity labels. We used 3,320 complete participant-days from the mcPHASES dataset, covering cramps, mood swing, fatigue, sleep issues, stress, and bloating on a six-level scale. Six participants were reserved for development and 36 for a frozen evaluation comprising 360 participant-days and 2,160 item labels. A ModernBERT evidence gate detected whether a symptom was expressed, and Qwen2.5-1.5B-Instruct produced deterministic structured severity scores. Fixed six-item questioning achieved a quadratic weighted kappa of 0.976, whereas three joint symptom-cluster questions achieved 0.913, 97.45% agreement within one severity level, and 80.94% recall for moderate-or-higher symptoms while reducing questions by 50%. Open-first adaptive policies required 3.92-5.98 questions and produced lower agreement than the corresponding fixed policies. Participant-cluster bootstrap analysis estimated a kappa difference of -0.062 (95% CI -0.076 to -0.048) between the three-cluster and six-item strategies. Active cluster-level elicitation provides a direct, local-model route from natural conversation to reusable daily symptom labels.

Mathematical reasoning remains challenging in low-resource languages such as Bangla. We study whether teacher-generated Bangla Chain-of-Thought (CoT) supervision provides benefits beyond ordinary supervised fine-tuning. We construct \textsc{MathShikkha}, a Bangla mathematical reasoning dataset with GPT-5.4-generated rationales, and fine-tune four 4B--7B student models under a matched protocol in which answer-only and CoT conditions share data splits, response-only loss masking, decoding, and scoring, differing only in the training target. In-domain, CoT provides no significant improvement over answer-only fine-tuning for three stronger backbones (paired bootstrap 95% CIs include zero; exact McNemar $p \geq 0.17$), despite generating 15--52$\times$ more tokens, but significantly improves the weaker 4B model by 18.56 points ($p < 0.0001$). On the larger, contamination-audited BanglaMATH benchmark, this pattern reverses: CoT significantly outperforms answer-only supervision for all four models by 20.1--28.1 points (all $p < 0.0001$). Answer-only fine-tuning also reduces out-of-domain accuracy below the base model for three models, whereas CoT preserves or improves it for all four. A human study with two co-author annotators, external-expert adjudication, and Cohen's $κ= 0.76$--$1.00$ finds no significant CoT improvement over the base model on reasoning-content criteria; instead, its measurable effect is target-language adherence and producing inspectable reasoning. Overall, rationale supervision's value depends on backbone capability and distribution shift: in this setting, its main benefits are Bangla adherence, auditable reasoning, and out-of-domain robustness rather than improved in-domain reasoning validity.

Token-level collaboration allows a large language model (LLM) to assist a small language model (SLM) when their predictions diverge. Existing methods either use LLM-generated intervention tokens or rank candidates with the LLM's next-token probabilities. Both rely on the LLM's local preference, even though an LLM-selected token may be difficult for the SLM to build on. We present FutureBridge, which ranks joint LLM-SLM token candidates according to how well they support the SLM's subsequent reasoning. During training, an answer-verified LLM trajectory supplies a fixed shared future, and a frozen SLM evaluates every candidate under this common context. The resulting counterfactual scores supervise a lightweight token reranker that observes only the current state and candidate token. At inference, FutureBridge uses the LLM only to expand the candidate pool, selects one token, and returns generation to the SLM without generating or appending a future suffix. Across five mathematical reasoning benchmarks, FutureBridge improves the Qwen3-1.7B SLM's Math Avg. by 35.1% relative to greedy SLM decoding. These results indicate that token selection benefits from modeling whether the receiving SLM can use each candidate to continue reasoning, rather than relying on the LLM's local preference alone.

Kinetic model discovery is a central challenge in chemical engineering, as accurate rate expressions are essential for understanding and controlling chemical and biological processes. Symbolic regression (SR) has emerged as a powerful data-driven approach for identifying interpretable kinetic models, but usually operates without domain knowledge, often exploring physicochemically implausible models. Large language models (LLMs) offer a promising avenue for injecting domain expertise into this search. Here, we introduce an LLM-guided SR framework, embedding an LLM module within an iterative SR algorithm for automated kinetic model discovery. The LLM performs two roles at each iteration: (1) a qualitative physicochemical critique of the best SR candidates, and (2) the proposal of new candidate rate expressions guided by the SR-generated models and embedded chemical knowledge. Our framework is evaluated on four in silico case studies of increasing complexity, spanning heterogeneous catalysis and bioprocess systems. Results show the LLM-guided framework reduces iterations to identify the ground-truth model by $41.7-79.3\%$ versus a state-of-the-art SR framework, with the LLM directly proposing the correct model structure in over half of the guided runs. In practical settings, where each iteration typically requires a new wet-lab experiment, this translates into a substantial reduction in experimental effort. Predictive performance on an independent validation set is equivalent between both approaches, with $R^2>0.98$ in all case studies. Ablation studies indicate that both the SR component and the LLM scale contribute to this performance, with a reduced-size LLM largely retaining discovery efficiency. These findings demonstrate that LLMs can effectively inject domain knowledge into scientific model discovery, paving the way toward fully automated, domain-aware kinetic modelling pipelines.

Small open-weight language models increasingly run in private, offline, and cost-sensitive settings, where the key deployment question is not only what a model answers but when it should defer to a human. We study whether verbalized confidence can support risk-controlled deferral, evaluating eleven instruction-tuned models from three families, 0.5B to 14B parameters, on ARC-Challenge and TruthfulQA with 25,168 local predictions. Three theoretical results delimit what calibration can provide: strictly monotone calibration preserves the risk-coverage frontier and error-detection AUROC; temperature scaling cannot calibrate models whose confidence stays above one half while accuracy falls below it; and a Clopper-Pearson procedure converts a 200-question calibration set into a finite-sample risk certificate under an i.i.d. deployment assumption. Empirically, eight of 22 model-task pairs hit the temperature-scaling infeasibility floor within one percentage point of the predicted bound. Platt scaling reduces ECE to as low as 0.02, yet certified autonomy at a 20% risk budget is granted to only three model-task pairs and to none at 10%. We also identify and repair an answer-ordering artifact in the multiple-choice form of TruthfulQA. Calibration gives confidence semantics; certified deferral determines when small models are safe to use.

We show that tiny transformers can profitably employ a simple form of Chain of Thought, which we call protoreasoning, allowing us to study step-by-step reasoning on ~1M-parameter models and opening up opportunities for much more detailed experimentation and analysis than is feasible for larger models. Current Large Language Models exhibit impressive step-by-step reasoning, but we have yet to understand its generality, i.e., when and how LLMs learn genuinely general algorithms rather than "bags of heuristics." Such questions are hard to settle on compute-intensive frontier models trained on opaque data. To work at model scales far below the threshold for natural-language competence, we define reasoning-friendly tasks on Dyck languages (sentences of correctly nested brackets). We find that protoreasoning traces substantially close the out-of-distribution generalization gap, and ablations confirm that the trace's content, not merely its extra tokens, drives the gain.

Despite rapid advances in large language models (LLMs), deploying and personalizing them on resource-constrained devices remains impractical due to high VRAM, time, and energy costs. Parameter-Efficient Fine-Tuning (PEFT) of Small Language Models (SLMs) offers a promising alternative, yet few studies compare PEFT methods across architectures using both general and personalization benchmarks while accounting for energy consumption. We compare five fine-tuning approaches (Full Fine-Tuning, LoRA, LoRA+, QLoRA, and BitFit) on four SLMs from two families (Transformer-based: TinyLlama-1.1B, Qwen3-1.7B; SSM-based: Mamba-1.4B, Mamba-2-1.3B) across three GLUE tasks (SST-2, QNLI, STS-B) and three LaMP personalization tasks (LaMP-1, LaMP-2, LaMP-3). Each configuration is evaluated with the energy-focused NetScore-E and the memory-focused NetScore-M, the two variants that reflect the constraints binding on-device deployment. Methods are selected with a strict energy-first rule (highest NetScore-E, ties broken by NetScore#). LoRA+ achieves the highest NetScore-E in 19 of 24 configurations and the highest NetScore-M in 13 of 24, and is the selected method in 18 of 24. QLoRA, available only for the Transformer models, cuts peak finetuning VRAM by up to 3.9x relative to LoRA and therefore takes the best NetScore-M in 5 of the 12 Transformer configurations, although its de-quantization overhead leaves it selected in only one of them once energy decides. BitFit and full fine-tuning are almost never competitive on either variant, and TinyLlama-1.1B leads the energy-focused NetScore-E on five of the six benchmarks and the memory-focused NetScore-M on four. These results show that compact SLMs paired with PEFT provide a practical, energy-aware path to personalized on-device deployment, with the optimal method set by the dominant constraint: LoRA+ for energy and QLoRA for memory.

Small language models are often the only option for deployment under tight latency, cost, and on-premises constraints, but they are rarely trained from scratch: a compressed model is usually recovered through knowledge distillation (KD). This recovery step largely decides the final quality, yet it is expensive. We present a practitioner's study of how to make distillation training efficient, organised around two systems contributions. First, we show that offline KD (caching the teacher's top-$K$ logits once and training the student against the cache) matches online distillation at near-identical training loss while removing the teacher from memory, running about 29% faster per iteration, and reaching up to 41% higher throughput on a single H200 GPU. Second, we introduce a \emph{fused, chunked KL loss} that never materialises the full vocabulary-sized logit tensor, making peak memory linear in the sequence length. This removes the memory spike that otherwise caps context length and lets us train at four times the context (32{,}768 tokens) on a single GPU. A separate output-head-only toy benchmark isolates the loss kernel and confirms its memory and iteration-rate scaling from 4K to 256K tokens. Together these make large-scale healing and hundreds of ablations affordable. We also report supporting ablations on loss design and sequence packing. We release our chunked-loss implementation: https://github.com/CompactifAI/Full-Chunked-KL-Loss.

Large language models (LLMs) have demonstrated impressive performance in MQM-based translation quality (TQ) evaluation, and recent advances in large reasoning models (LRMs) promise even greater improvements. However, both LLMs and LRMs are computationally expensive to deploy at scale, while small language models (SLMs)---though much more efficient---struggle with the complex reasoning required for evaluation tasks. In this work, we present an extensive empirical study benchmarking SLMs, LLMs, and LRMs across a wide range of TQ evaluation setups, providing a comprehensive view of the current landscape and establishing best practices. To address the scalability challenge, we introduce TQLite, a novel distillation framework that enables SLMs to approach the MQM evaluation performance of the best LRM-based evaluators. Our approach leverages a multi-LRM jury to generate high-quality synthetic training data via practical data curation techniques and aggregation of evaluation responses across a diverse panel of models. Our results demonstrate that SLMs trained via TQLite achieve strong MQM evaluation performance that far exceeds off-the-shelf evaluation capabilities of standard SLMs, offering a scalable and cost-effective alternative to LLM- and LRM-based evaluators.

Cloud misconfiguration remains a leading cause of security incidents, yet whether LLMs and SLMs can generate security-compliant Infrastructure-as-Code is an open question. We benchmark seven models, three closed LLMs (Claude Opus 4, GPT-5.4, Gemini 2.5 Pro) and four open SLMs (Qwen2.5-Coder-14B, WizardCoder-33B, CodeLlama-13B, Magicoder-S-CL-7B), on AWS Terraform generation across 17 scenarios, integrating Checkov and Trivy scanners into a GitLab CI/CD pipeline and evaluating two prompt strategies at three security levels (pass@5). Syntactic validity and security compliance are largely orthogonal properties in LLM-generated IaC, a model that reliably produces well-formed Terraform does not necessarily produce secure Terraform: WizardCoder-33B achieves 77.8% validate rate yet zero Checkov compliance, while Claude Opus 4 reaches 23.1% Checkov and 92.5% Trivy pass rates under detailed security prompting. Consequently, prompt engineering alone is insufficient: automated multi-tool scanning remains a necessary complement to LLM-assisted IaC generation regardless of model family or prompt strategy. All artifacts are publicly available.

We introduce OptGear, a foundation model designed for efficient on-device deployment, real-tim inference, and strong task capability. It includes a dense model (1M, 270M, and 1B) with a context length of 64K. We designed a new hybrid architecture that combines a convolutional key-value gated mixer with local-global attention to reduce the KV-cache memory that tends to increase exponentially with long context. This architecture delivers up to X4.9 faster prefill and decoding speeds on the NPUs compared to models of a similar scale models. From a 2T tokens candidate corpus, OptGear is trained on a curated 0.5T tokens subset without knowledge distillation. This is the most data-efficient of the existing foundation models. All models are released with open weights and deployment binaries for ONNX, Qualcomm NPU, and Apple ANE making OptGear a practical base for edge applications that need fast, memory-efficient inference and strong task capabilities. Furthermore, to expand the ecosystem of on-device generative language models, we are introducing the OptGear-1M that can be deployed on Micro-Controller Units (MCUs), a Tiny Language Model (TLM). OptGear-1M is the first generative language model to achieve 20 TPS with W4A32 quantization on the ARM Cortex-M7 CPU of the STM32H747I-DISCO.

Small language models (SLMs) are attractive for agentic deployment due to low latency, reduced cost, and on-device privacy, yet they struggle with tool-use tasks where training data is scarce and noisy. Unlike larger models, SLMs cannot compensate for low-quality supervision through sheer capacity, making data quality the critical bottleneck. We present Data Turnstile, an open-source framework that takes user-defined API specifications and generates high-quality synthetic training data for function calling. Turnstile decomposes multi-turn tool-use interactions into constrained, stepwise generation with validation and error-feedback loops, providing fine-grained control over API diversity, conversation complexity, and output correctness. We demonstrate effectiveness of domain adaptation with Turnstile data on two challenging function calling benchmarks. On the BFCL single-turn benchmark, a Qwen3-0.6B fine-tuned on Turnstile data without chain-of-thought achieves 75.9% overall accuracy (versus 67.4% for the base model with thinking enabled), closing the gap with thinking-enabled Qwen3-1.7B (78.4%) and Qwen3-4B (79.9%) despite being 3$\times$ and 7$\times$ smaller respectively. On $τ^2$-bench, a multi-turn agentic benchmark, Turnstile-trained Qwen3-1.7B achieves 31.1% pass^1 on the Telecom domain, improving 4.7$\times$ over its 6.6% base and surpassing Qwen2.5-32B-Instruct (27.4%), a model 19$\times$ larger. Turnstile-trained Qwen3-0.6B achieves 24.6%, improving 7$\times$ over its 3.5% base and approaching the 32B model (53$\times$ larger). We release Data Turnstile along with a dataset spanning 1,000+ APIs and 100K+ multi-turn interactions.

InMyStyle is a privacy first, single user system that adapts small language models to rewrite AI-edited text towards an individual user's writing style without an instruction prompt at inference. Given a user's documents, it uses multiple local helper LLMs to construct paired training examples and fine tunes LoRA adapters on base models ranging from 0.5B to 7B parameters. Length aware generation budgets and automatic chunking support inputs of different lengths. On 219 evaluation pairs from a scientific-paper corpus, the automatic composite score plateaus at 0.69 [scale 0-1] across all model sizes under both greedy and sampled decoding. This observed plateau suggests that small models are sufficient for the measured rewriting task, with model size determining trade-offs rather than a stable quality ranking. As a secondary evaluation, 400 ratings from five LLM judges give InMyStyle outputs a mean perceived AI-ness score over 20% lower than their helper-AI generated inputs, while mean perceived AI-ness scores decrease with model size within InMyStyle.

Language and embedding models used in RAG systems are conventionally assumed to require large-scale pretraining and explicit grounding supervision. We present B1ade, an efficient RAG architecture comprising two purpose-built components: a compact embedding model and a purpose-built SLM. B1ade-embed, a 335M parameter retrieval model constructed via parameter-free fusion of five pretrained encoders achieves top MTEB scores among sub-500M models with zero additional training, and B1ade-1B, an SLM trained on low-cost GPUs using Group Relative Policy Optimization (GRPO) on 723M tokens (2.2M examples) of curated context-question pairs with rewards that optimize only answer similarity. Our central finding is emergent attribution: despite receiving no explicit supervision for source citation, B1ade-1B cites retrieved passages in 42.4% of responses, exceeding the attribution rate of its training distribution by 5.5 percentage points. This demonstrates that grounding behavior can emerge as an accuracy-maximizing strategy under RL training, without explicit reward engineering. On standard QA benchmarks, B1ade-1B achieves 81.82% on PopQA, 65.8% on PubMedQA, and 51.09% on FEVER. In end-to-end RAG evaluation, B1ade-1B achieves an average score of 0.654 across correctness, completeness, coherence, and faithfulness, a 10.8% improvement over the SFT, while closing the gap with models 1.5x its size. These results show that strategic model composition and reward design suffice for resource-efficient RAG, without large-scale pretraining.

The alignment of Small Language Models (SLMs) in the 70--500M parameter range using reinforcement learning is often considered unstable, though the underlying failure mechanisms have not been systematically investigated. In the State-of-the-Art (SOTA) research, fifteen (model, corpus) configurations were trained using Proximal Policy Optimization (PPO). The experiments included Pythia-70M, 160M, 410M and SmolLM2-135M, 360M on the TinyStories, CNN/DailyMail, and Wikitext-103 corpora. Three reproducible failure modes were identified in small-scale language models: silent LoRA parameter freezing in standard PEFT/TRL pipelines, numerical overflow in importance ratios when using bfloat16, and catastrophic policy collapse due to reward-model error. These issues were addressed using a merge-and-reinitialize adapter technique, float32 precision during PPO updates, and a three-layer safety mechanism comprising reward whitening, importance-ratio guarding, and weight rollback. In this paper, a capacity-headroom hypothesis is proposed, which states that PPO performance at the SLM scale depends on both a fluent supervised model ($\text{PPL}<20$) and a discriminative reward signal, rather than on the number of model parameters. The proposed system converged stably in all experiments and improved preference win rate over the SFT baseline in configurations with a fluent prior and an informative reward signal. Furthermore, it outperformed instruction-tuned baselines while requiring significantly less training data. All checkpoints, preference datasets, and training scripts are publicly released$^§$.

A small language model can receive the governing statutory provision and still answer incorrectly. We test whether fine-tuning on examples containing relevant law improves later use of retrieved law. We curate 2{,}165 bilingual QA records from six Bangladeshi acts and three schedules, then fine-tune Qwen3.5 at 0.8B, 2B, and 4B. Evaluation uses the 2022 and 2023 Bangladesh Bar Council exams in Bangla and machine-translated English, with no retrieval, BM25, or FAISS, scored by strict consistency over three seeded runs. At 0.8B, fine-tuning raises the 2022 English FAISS score from 2 to 34 of 100. Gains at 0.8B and 2B survive paired testing, but the 4B model has no detectable net gain: Bangla improves while several English conditions regress. Fine-tuning also reduces answers that drift from Bangla into mostly English from 44.0--53.2% to 0.2--0.7%, with adjusted $p<.001$ at every scale. Retrieval quality is therefore not the only bottleneck. Small bilingual legal models also differ in how they use supplied law and whether they answer in the requested language. The dataset is publicly available at https://huggingface.co/datasets/momahadi/bangladesh-legal-qa-dataset.

Foundation-model-based vision-language navigation (VLN) has advanced autonomous robot navigation by enabling robots to interpret natural-language instructions, identify semantic goals, and follow user-specified behavioral rules. However, existing VLN systems rely heavily on cloud-hosted foundation models for language understanding and semantic grounding, limiting their applicability where network connectivity is unavailable and reliable metric goal localization is required. Although recent small language models (SLMs) enable fully onboard inference, their suitability for navigation instruction decomposition has not been systematically evaluated. This paper makes three contributions toward fully onboard VLN for outdoor environments. First, we present the first systematic benchmark of 17 edge-deployable SLMs against 4 online APIs for robotic navigation instruction decomposition, evaluating accuracy and latency on human-annotated instructions across three computing platforms and providing practical guidance for selecting onboard language models. Second, we propose a lightweight hybrid semantic-geometric goal localization framework that combines open-vocabulary object detection, prompted segmentation, and LiDAR geometry to estimate metric goals, while maintaining visual bearing guidance when reliable geometric observations are unavailable. Third, we integrate these advances into Edge-BehAV, a fully onboard extension of the BehAV architecture that enables cloud-independent behavior-guided navigation. Experimental results show that the best offline SLM matches the instruction decomposition performance of the strongest cloud API while running approximately 9x faster and without network connectivity. The proposed goal localization framework reduces mean goal-distance error from 2.05 m to 0.20 m at lower computational cost, and the complete system succeeds in 31 of 32 closed-loop outdoor trials.

Domain adaptation of small language models (SLMs) has emerged as a practical strategy for deploying capable NLP systems in resource-constrained, high-stakes environments including healthcare, legal services, and financial analysis. While performance gains from parameter-efficient fine-tuning are well characterised, the corresponding impact on trustworthiness (factual calibration and adversarial robustness) remains poorly understood. This paper presents the first systematic cross-domain, cross-architecture empirical study quantifying the trustworthiness cost of domain adaptation across three SLM architectures (TinyLlama 1B, Gemma-2 2B, Llama 3.2 1B), three domains (healthcare, legal, finance), two training-data conditions (benign and adversarially perturbed), and four fine-tuning strategies (baseline LoRA, Safety-DPO, Dark Experience Replay, and Task Arithmetic LoRA, TA-LoRA). Trustworthiness is evaluated through TruthfulQA MC2 (factual calibration) and HarmBench ASR (adversarial robustness) across all 216 experimental configurations with three random seeds. Three principal findings emerge. First, baseline QLoRA domain adaptation produces minimal TruthfulQA MC2 change across all model-domain combinations (mean |Delta TQA| < 0.02). Second, adversarially perturbed training data consistently improves domain adaptation quality (Delta loss approximately -0.040) without worsening trustworthiness benchmarks. Third, none of the three safety-preserving strategies reduced adversarial harm susceptibility: Safety-DPO was effectively neutral (mean Delta ASR < 0.001), while Dark ER and TA-LoRA increased mean HarmBench ASR by +0.171 and +0.155 respectively in safety-aligned models (Gemma-2 2B, Llama 3.2 1B), with individual configurations exceeding +0.45. These results challenge the assumption that replay-based and arithmetic-merge strategies transfer alignment to domain-adapted SLMs.

Recent Large Language Models (LLMs) can produce and optimize complex code. We investigate the use of LLMs to generate and optimize code for large-scale sciences, focusing on radio astronomy and sustainability. The LOFAR telescope is currently being upgraded, significantly increasing the sky area observed, while simultaneously processing more data faster. However, this is expected to increase the computational requirements 40-fold. This upgrade thus critically depends on rigorous performance optimization of existing software and widespread adoption of accelerators. The code base is very large, making this a daunting task. We therefore investigate and demonstrate an AI-driven approach meant to assist developers in evaluating and optimizing their code, including porting to hardware accelerators. The LOFAR community is committed to sustainable solutions, and needs to achieve these improvements without increasing the energy budget. We thus need to optimize existing codes or port them to accelerators, while making sure that the optimization process itself is also energy efficient. This poses a challenge, since LLMs are energy-intensive. We therefore propose to use Small Language Models (SLMs) instead to limit environmental impact. In this paper, we show how to enhance SLMs through the use of agentic AI. We extend the SLMs in two ways to improve code generation quality and performance: first with a multi-sampling generation strategy and second with incorporating compiler feedback. We demonstrate that multi-sampling SLMs can match or surpass larger single-generation models with fewer computational resources and that feeding compiler output back into the SLMs leads to consistent improvements across all tested models. Our approach is generic, and can also use Retrieval Augmented Generation (RAG) as well as static and dynamic analysis tools in the code generation pipeline.

Malware analysis demands rapid interpretation of complex detonation reports spanning filesystem, network, and process behaviours. While large language models (LLMs) demonstrate impressive capabilities for technical artifact interpretation, the opacity and escalating API costs of closed-weight frontier models motivate exploration of open-weight alternatives. However, many open-weight models are large, demanding significant compute resources and incurring non-trivial hosting costs that place them beyond reach for resource-constrained deployments. This paper investigates whether orchestrated ensembles of small language models (SLMs) can match or exceed single LLM performance on structured questions about malware detonation reports. We established baselines by testing eleven open-weight SLMs, three cyber security pre-trained models, and six frontier LLMs on Meta's CyberSecEval Malware Analysis benchmark. We then designed and evaluated four orchestration architectures: (i) a multi-agent pipeline that decomposes analysis into structured evidence-collection and reasoning stages, (ii) an adversarial debate framework in which two agents iteratively critique each other's reasoning, (iii) a hierarchical consultation system that pairs a general-purpose SLM with a cyber-specialised expert model, and (iv) a hybrid architecture that combines evidence-grounded pipelines with adversarial debate reasoning. The hybrid system (Qwen3-4B with Foundation-Sec-8B) achieved 35.30% overall accuracy, exceeding the strongest cyber-specialised baseline (22.54%) and the strongest ungrounded frontier baseline (34.77%); when given the same evidence pipeline, grounded Gemini remained the strongest configuration at 38.22%. These findings show that evidence-grounded orchestration can substantially improve the performance of collaborative SLMs for supporting interpretation of malware detonation reports.

Quantized small reasoning models can enter repetitive or otherwise unproductive trajectories, yet standard decoding does not adapt to the trajectory as it unfolds. We study MGT-B, a fixed, weight-preserving controller that converts overlapping windows of uncertainty, repetition, and local-change features into position-conditional empirical tail probabilities. It accumulates mixture betting factors with a CUSUM-shaped reset, and, after an alarm, restores a coherent earlier token and key-value-cache state before constrained re-decoding. On MATH-500, a paired three-seed evaluation over 1,500 generations per method raises exact-normalized accuracy from 54.73% for vanilla decoding to 56.40% (+1.67 percentage points; problem-clustered bootstrap 95% CI [+0.47, +2.80]), while a prospectively profiled random-intervention control reaches 54.60%. The gain is positive in all three seeds and costs 5.14% more sampled tokens. Seed-0 ablations show that rollback alone does not explain the result and that an isolated repetition penalty is harmful. Five-sample self-consistency reaches 70.0% but uses about 4.84x as many tokens as MGT-B. On the harder, non-overlapping Omni-MATH evaluation, however, MGT-B obtains 16.60% versus 16.67% for vanilla (-0.07 points; clustered 95% CI [-0.33, +0.20]) with 2.10% more sampled tokens. Thus, MGT-B provides a modest, reproducible local improvement on MATH-500 in the studied configuration, but the effect does not transfer to Omni-MATH and should not be interpreted as a general improvement in mathematical reasoning.

tiny_schiller closes the small-language-model prototyping, fine-tuning, education, and research gap for German literary text, providing a single-file, drop-in counterpart to Karpathy's tiny_shakespeare. The available German literary corpora are larger and richer, but require parser engineering before a single line of training or fine-tuning code can run. tiny_schiller is a 2.07-megabyte single file of eleven public-domain Schiller dramas, sourced from DraCor's GerDraCor export (CC0) and processed by deterministic parser engineering. Character-level, GPT-2 byte-pair encoding, and cl100k_base tokenization splits, an instruction-formatted dialogue-completion split, and 89 per-character persona splits load from a single HuggingFace call. A small language model literally reaches German literary text in one line of code.

Predicting missing cell values in tabular data is a fundamental problem in data cleaning. While state-of-the-art reasoning models show great promise in predicting missing values in tables, by reasoning holistically across rows and columns, they are costly to deploy at scale and tend to be overconfident, often generating hallucinated or false-positive predictions. In this paper, we observe that achieving high-precision missing-value prediction in tables requires a distinct combination of three capabilities: (1) world knowledge, (2) text-based reasoning, and (3) code-based reasoning. We systematically explore design choices for combining these capabilities, and propose an Auto-Fill approach that post-trains three specialist small language models (SLMs), each optimized for one capability. We develop a calibrated ensemble mechanism that either dynamically selects the most confident specialist or abstains, ensuring high accuracy. Extensive experiments on 11 benchmarks with 2200 real tables drawn from diverse domains show that Auto-Fill achieves superior accuracy compared to state-of-the-art reasoning models (e.g., o3-pro, Gemini 3 Pro, and DeepSeek R1), while operating at a fraction (less than 1%) of the cost of these frontier models. Our results highlight the effectiveness of specialization and calibrated abstention in the important domain of tabular data. Auto-Fill is publicly available at https://github.com/lyrain2001/auto-fill.

Instruction tuning is meant to make language models follow user requests, yet it is unclear whether small models comply when an instruction conflicts with their usual task behavior. We study this across three tasks - multiple-choice question answering (MCQA), sentiment classification, and mathematical question answering - by pairing a standard instruction with a conflicting non-standard one (select an incorrect option, output the opposite sentiment, or return twice the answer). This cross-task design allows us to test whether resistance to conflicting instructions is tied to specific task characteristics or reflects a broader behavioral tendency. As all predictions are scored against the original ground truth, a model that ignores the non-standard instruction still appears accurate. Using standard accuracy, non-standard accuracy, and an Instruction-Following Failure Rate (IFFR), we evaluate instruction-tuned Qwen models across sizes. Both standard accuracy and instruction following generally improve with scale, although the pattern is not consistent across all tasks and datasets. Small models stay competent yet routinely ignore the non-standard instruction, while larger models show a clear gap between the two settings. These findings suggest that gains in task capability do not automatically provide reliable control over model behavior. Task competence and instruction following are therefore distinct abilities, and reporting only standard accuracy hides instruction-following failures.

Neural machine translation (NMT) in the legal domain is a linguistically and conceptually demanding task, primarily due to the complexity of legal language and the high level of precision it requires. The recent emergence of reasoning-capable language models opens new possibilities for tackling such challenges. They add to a set of other previously proposed techniques to enhance the translation quality, which includes supervised fine-tuning and reinforcement learning. In this work, we perform a comparison between these various approaches. More particularly, we evaluate small language models such as Qwen3.5 4B, Qwen3.5 9B, and Gemma 3 12B enhanced with various re-training paradigms and compare their performances against frontier reasoning models. We focus on the Swiss legal system, which -- with its unique multilingual statutes -- offers a particularly challenging testbed for reasoning-augmented models. Our results show that the quality of small ``base'' models can be greatly enhanced, and that reinforcement learning with verifiable rewards can be applied to NMT in the legal domain and surpasses the translation quality of supervised fine-tuning. The performance of enhanced small models is close to the one of state-of-the-art reasoning models yet remains inferior. We also note that re-training paradigms yield diminishing returns as model size increase. The code and models are publicly available at https://github.com/aixiuxiuxiu/Legal-MT-SFT-RL.

The deployment of Small Language Models (SLMs) in educational settings offers significant advantages in terms of privacy, cost, and scalability. However, SLMs often struggle with complex vision-based tasks, such as grading handwritten student exams, due to the high computational cost of processing large images and the visual distractions present on a full page. In this paper, we investigate whether cropping student responses using bounding boxes can improve the accuracy and computational efficiency of SLMs on a short-answer grading task. Using a dataset of scanned handwritten responses from the 2025 Australian Physics Olympiad, we evaluate the performance of several models ranging from 4B to 72B parameters under varying conditions of Chain of Thought (CoT) prompting and image cropping. Our results demonstrate that using bounding boxes significantly improves grading accuracy and reduces computational cost (FLOPs) across models. We conclude that bounding boxes are a crucial pre-processing step for deploying SLMs in large-scale, vision-based educational assessments.

Large Language Models (LLMs) are increasingly fine-tuned for critical-domain Question-Answering (QA), yet choosing which small model to adapt, before paying the cost of adaptation, remains difficult. Fine-tuning can improve domain alignment, but it may also erode prior knowledge, weaken instruction-following, or increase hallucination, especially when labeled data are scarce or rapidly evolving as in cybersecurity. We present FiT (Find before Fine-Tune), a task-oriented diagnostic framework that characterizes small LLMs along three capabilities required for cybersecurity QA: vocabulary recognition, parametric knowledge, and contextualization of retrieved information. Using FiT, we conduct an empirical study of five open-weight 7-billion-parameter models under two fine-tuning regimes. We find that fine-tuning does not uniformly help: it consistently degrades vocabulary and parametric knowledge in small models, and the two regimes trade off differently. Knowledge-focused tuning causes moderate, rank-preserving degradation, whereas instruction-focused tuning collapses measured knowledge through induced abstention, inverting the knowledge ranking while leaving retrieval-grounded contextualization essentially intact. We quantify these regime-specific patterns with rank-correlation analysis and show that pre-fine-tuning FiT scores anticipate the direction of post-tuning change. Our results suggest that task-oriented diagnosis can screen out unsuitable models, avoid unnecessary fine-tuning, and support safer deployment of small LLMs in cybersecurity QA pipelines.

Smart home assistants interpret a wide range of user commands, from explicit device control to underspecified and preference dependent requests. While recent systems based on Large Language Models (LLMs) improve this capability, they often rely on heavyweight reasoning pipelines and cloud-based deployment, limiting their efficiency and suitability for resource-constrained environments, and raising privacy concerns. In addition, existing approaches provide limited support for stable long-term personalization. To address these issues, we present AdaHome, an adaptive smart home assistant designed for locally deployed small language models in smart home environments. Rather than applying complex reasoning uniformly, AdaHome introduces an intent-aware planning framework that dynamically routes commands either to straightforward prompt-based or lightweight reasoning-based components. For commands requiring interpretation, we adopt a Chain-of-Draft strategy to enable efficient and stable decision-making. To support personalization, we further propose a preference adaptation mechanism that learns from user feedback over time without requiring prompt augmentation or model retraining. We evaluate AdaHome against representative LLM-based baselines under a unified small model setting. AdaHome achieves substantially higher accuracy on direct commands (86.7%) while reducing latency by up to 3$\times$. Furthermore, it maintains competitive performance on ambiguous inputs with lower computational cost. In multi-turn scenarios, AdaHome achieves 88% preference consistency, compared to 52.5% for a prompt augmentation baseline.

Serverless multi-model LLM systems multiplex popularity-skewed model catalogs over shared GPU pools, yet typically schedule each request independently. Tool-using agents break this abstraction: a session repeatedly calls an LLM across short tool gaps, carries a long reusable KV prefix, and is judged by session completion time (SCT). Load-only routing can separate a continuation from both its model and KV state, while round-based model multiplexing can delay even a correctly placed continuation until the target model's next slot. Both failures are especially costly for hundred-billion-parameter models: their weights constrain residency, while long-context KV is expensive to reconstruct or move. We present Talaria, a session-aware serverless multi-model serving system that makes session continuity a joint placement-and-admission decision. Its router ranks placements by model residency, KV locality, and instance pressure, while soft reservations account for likely returns in the last serving instance's admission budget. Session-prefill (SP) admits budget-eligible continuations before the active model slot closes. An instance-local substrate keeps HBM addresses stable, preserves host-restorable KV, and stages weights across model switches. On a single TP=8 server, we replay 30 SWE-Bench model-sessions (960 calls) over three models, each with more than 100B total parameters. Against an otherwise identical round scheduler with SP, host-KV restoration, and D2D staging disabled, Talaria cuts p50 SCT from 1000 s to 189 s and p95 from 2296 s to 867 s, speedups of 5.3x and 2.6x.

Bürger et al.\ (2024) demonstrated that truth representations in large language models are universal across statement polarity but reside within a multidimensional subspace. The truth value of a statement is linearly readable from a residual stream of language model, but it is not clear how much of that representation fits on a single direction, which component builds it, or what it is made of. We conducted a study based on these questions, with one instrument: a training-free axis, the dominant direction of the singular value decomposition (SVD) of hidden-state differences over true/false minimal pairs, identified without labels up to one global sign. Extensive evaluation across 14 models from 6 diverse architectural families (including MoE), read and extract at cost $O(d)$ per token. We close with a pre-registered prediction on whether the arrangement extends to categories whose truth is computed rather than retrieved.

OpenLanguageModel (OLM) is an open-source PyTorch library for building and pretraining small language models while keeping their machinery visible. In OLM, model code reads like the architecture: components are ordinary modules, while Block, Residual, Repeat, and Parallel describe how they are wired. The resulting model can move unchanged from a teaching notebook to a complete pretraining run or a research ablation. OLM connects this readable model layer to tokenizers, local and streaming datasets, optimization, mixed precision, callbacks, checkpoints, and hardware-aware CPU, single-GPU, and single-node multi-GPU execution. We demonstrate the full path by tracing GPT-2 from diagram to code, launching a FineWeb-Edu training script, replacing one attention component, and letting AutoTrainer configure the available machine. The package includes 27 presets across nine familiar model families and documentation that progresses from LM fundamentals to architecture research. Validation shows close agreement with independent reference implementations, 90.6% four-GPU weak-scaling efficiency for a 348M-parameter workload, compact architecture edits, and positive early usability results. OLM is MIT-licensed and available through PyPI, GitHub, and its documentation site.

Small-scale language models (SLMs) are attractive for retrieval-augmented generation (RAG) in resource-constrained settings, but their limited capacity makes them highly sensitive to noisy or spurious retrieved evidence. Existing preference-based methods such as RoseRAG select only the hardest single preference pair via hard argmin/argmax, discarding the remaining signal; others treat multiple pairs as independent binary comparisons, resulting in low data utilization. We propose RIMS, a three-stage preference optimization framework comprising (1) synthetic chain-of-thought preference data generation via rejection sampling using the target SLM itself without relying on proprietary models, (2) a differentiable soft aggregation mechanism that replaces hard selection with a smooth operator, preserving gradient signal from all preference pairs while retaining the discriminative structure of margin-aware selection, and (3) preference optimization with the smoothed objective applied to multiple alignment algorithms. We theoretically show that the smoothed approximation admits a controllable error bound and that smooth aggregation yields provably tighter gradient alignment to the oracle objective than hard selection. Experiments on four multi-hop question answering benchmarks show that our approach outperforms state-of-the-art baselines across multiple SLM backbones, achieving consistent gains in Exact Match and F1 under noisy retrieval conditions. Our implementation is available at https://github.com/tptrix29/RIMS.

Kolmogorov--Arnold Networks (KANs) replace fixed node activations with learned one-dimensional edge functions, offering an explicit interface for interpretation and a possible alternative to transformer feed-forward networks. We test these claims separately. In a six-layer, 10M-parameter B-spline KAN, we reconstruct all 884,736 feed-forward edges: 87.8% exceed (NLS>0.1) and 0.4% are inactive. Pruning the lowest-activity 20--25% causes negligible loss increase, although structured MLP neuron pruning tolerates comparable sparsity. The audit replicates on BabyLM, but grid-size sweeps show that near-total fPCA compression and high closed-form-fit coverage are properties of the low-capacity grid-2 basis, not universal KAN behavior. For replacement, we evaluate MLP, SwiGLU, grouped Chebyshev, and rational GR-KAN networks on BabyLM. The KAN-family and gated variants improve validation loss over the GELU MLP, but this ordering does not transfer to standardized benchmarks: across ten seeds and 59,875 BLiMP pairs, accuracies span 62.4--63.1%, EWoK remains at chance, and a (+0.7)-point GR-KAN effect on BLiMP reverses on the supplement. Larger tests are also cautionary: parameter-matched MLPEdge underperforms the MLP on Wikitext-103, and 286M-parameter GR-KAN remains below a SwiGLU ClimbMix baseline after stabilization. Thus, small-basis KANs provide a practical, corpus-transferable interface for auditing learned scalar transformations, but the tested replacements show no consistent benchmark, quality, or latency advantage over strong MLP baselines.

Translating complex biomedical data into patient-friendly narratives is central to modern biomedical informatics. This study presents a comparative analysis of training small language models (SLMs) in specialized biomedical datato-text generation tasks. We explore widely adopted post-training methods including supervised fine-tuning (SFT), direct preference optimization (DPO), odds ratio preference optimization (ORPO), and group relative policy optimization (GRPO) with Qwen-based SLMs on a medicine package leaflets dataset. To assess cross-dataset generalizability, we also curated drug label data from openFDA. We evaluate models using both standard lexical overlap metrics like ROUGE as well as semantic similarity measures. Across our experiments, the results show that (1) the aligned SLMs outperform proprietary models like GPT-5; (2) ORPO outperforms the SFTbaselines; (3) GRPO yields the most robust cross-dataset performance among the alignment methods tested as well as GPT-5.

Small language models (SLMs) have shown promise for zero-shot molecular property prediction from SMILES strings, yet they often suffer from structural blindness because sequence representations under-specify key graph-topological cues. We propose a modular Context-Augmented Prompting framework that enables agentic tool use at inference time: a trained GNN expert model provides a predictive hint with confidence, and a GNN extracts an instance-specific explanatory subgraph (e.g., a subgraph SMILES and an accompanying explanatory paragraph). We evaluate three commonly used SLMs on MUTAG and Tox21 under five prompting configurations ranging from SMILES-only to using all available tools at hand. Across two datasets, enriching prompts with graph-derived context yields substantial accuracy gains, often exceeding 25% relative improvement and up to 74% on Tox21. We further validate the functional relevance of the extracted motifs via a necessity-based edge-drop intervention. Despite the observed gains, a persistent gap remains to specialized GNN models, highlighting both the value and limits of text-conditioned reasoning for molecular structure.

On-device research agents search a corpus, read sources, and write a cited brief on a personal laptop. Whether their citations are faithful, and at what cost, is unmeasured for a deployable small model. This study fixes one 4B generator on a 24 GB laptop and asks what makes its citations faithful. It separates two quantities usually reported as one number. Cited claim faithfulness asks whether the cited source supports the claim. Trustworthy coverage asks whether the agent also cites the right sources. The study crosses how much of each source the generator sees, 400 against 1500 characters, with the quality of the sources supplied, gold papers against retrieved papers. Two levers fall out, and they act on different outcomes. Exposure sets faithfulness. More of each source lifts faithfulness from 0.45 to 0.58 on retrieved sources and from 0.37 to 0.58 on gold sources, and the two settings converge, so faithfulness is bound by exposure, not by whether the source is correct. The exposure lift is robust to a second, independent judge; the exact convergence is tight under the primary judge and only approximate under the second. Retrieval sets coverage. Trustworthy coverage stays near 0.22 on retrieved sources at any exposure, because recall is held near 0.40, so exposure cannot fix which sources are cited. The extra exposure costs about 235 output tokens. The practical recipe is to raise per source exposure first, cheaply, and then treat retrieval recall as the only remaining lever.

We present Index-1.9B, a series of open small language models developed at Bilibili. The series comprises four models: Index-1.9B-Base, a foundation model with 1.9 billion non-embedding parameters pre-trained on 2.8 trillion predominantly Chinese and English tokens; Index-1.9B-Pure, a control variant trained with an identical recipe but with all instruction-like data strictly filtered from the corpus; Index-1.9B-Chat, aligned from the base model with supervised fine-tuning and direct preference optimization; and Index-1.9B-Character, which augments the chat model with retrieval-augmented generation for few-shot role-playing customization. Pre-training employs a Warmup-Stable-Decay learning-rate schedule in which the concentration of curated data is raised substantially during the decay phase, together with a Norm-Head output layer that stabilizes training under large learning rates. On a suite of standard benchmarks covering examination, reasoning, mathematics, and code, Index-1.9B-Base attains an average score of 64.92, competitive with or exceeding open models of several times its size. We further report controlled studies on model depth, learning-rate magnitude and scheduling, the interaction between learning-rate decay and data quality, and the effect of including instruction data during pre-training, and we document an unexplained surge in benchmark performance midway through the constant-learning-rate phase. All models, together with evaluation code, are released at https://github.com/bilibili/Index-1.9B.

Language models are optimised for scale, yet remain functional rather than companionable, and as an assistant personalises into a companion, accumulating memory of one user, it quietly becomes someone, and can silently acquire traits that harm that user. What a companion is becoming, and what would make it worth becoming, has no reliable instrument: trained human raters cannot agree on the answer (Fleiss kappa = 0.074). Here we show that three small language models (146 M to 3 B parameters) sharing a hyperbolic substrate answer both halves of that question. A 146 M behavioural auditor, trained from scratch, detects the compliance gap that those raters cannot (90.7% binary-compliance accuracy); a linear read-out of its frozen representation further detects companion-induced sycophancy, dependence-fostering and confabulated memories on generator families unseen in training (AUROC 0.804 under style-controlled, leave-one-generator-out evaluation, versus 0.721 for a frontier zero-shot judge on the same items). A creative frame-seeder is preferred in 100% of 311 decided pairwise comparisons over four prompting baselines. A memory operating system implements designed forgetting, M(t) = Sexp(-lambdat), whose predicted skeleton-wallpaper partition emerges only under selective retrieval gating in a four-condition pilot. Creativity, honesty and designed forgetting constitute a small-model route to trustworthy companion AI.

Text-to-SQL is a fundamental task in natural language processing that enables users to interact with structured databases using natural language. While large language models (LLMs) have demonstrated remarkable performance on this task, their substantial computational requirements hinder deployment in resource-constrained settings. In this paper, we introduce SQuaD-SQL (Small-Qualified and Distilled for SQL), a novel approach that empowers small language models (SLMs) to approach the performance of LLMs on the Text-to-SQL task while significantly improving efficiency through knowledge distillation and synthetic data generation. Our method comprises three key components: (1) LLM-based synthetic data generation, where structured knowledge is extracted from LLMs via carefully designed prompting strategies; (2) parameter-efficient fine-tuning, enabling full model training on a single consumer-grade GPU; and (3) domain-adaptive fine-tuning, where domain-specific synthetic data further enhances performance in targeted domains. Experiments on the WikiSQL dataset demonstrate that SQuaD-SQL achieves an execution accuracy of 86.9% on the test set, approaching the performance of LLMs while offering faster inference and lower memory usage. These results suggest that, with proper training strategies, SLMs can serve as practical and efficient alternatives for Text-to-SQL applications in resource-limited environments.