On Predicting Vulnerability Severity Using In-Context Learning: An Industrial Case Study

Modern software systems require earlier and more scalable vulnerability severity assessment to reduce exposure to high-impact security flaws. Security analysts typically assign CVSS scores, but this manual triage does not scale with the growth of disclosed vulnerabilities and often depends on cloud LLM services that raise confidentiality concerns. This paper presents an industrial case study on predicting CVSS v3.1 scores directly from vulnerable C/C++ snippets using in-context learning with locally deployable, open-source LLMs. We compare proprietary data with the Big-Vul dataset, showing sufficiently aligned CVSS distributions to justify Big-Vul as a proxy for industrial data when constructing prompt-based testbeds. We then vary in-context configurations and model parameters, evaluating CodeLlama2-7B, CodeLlama2-13B, Mistral-7B, gpt-oss, and GPT4o-mini using mean squared error (MSE) and feasibility metrics. Our results show that medium-sized open-source code models, particularly CodeLlama2-7B, can approximate the best cloud performance for CVSS regression when guided by lightweight, output-constraining prompts, offering a practical, privacy-preserving building block for severity triage in industrial settings.

A Unified Model for Cross-Domain Clone Detection via Model Merging

The growing diversity of code clone types, from syntactic copies to cross-language semantic clones to AI-generated duplicates, has created a fragmentation crisis in clone detection. Current deep learning detectors are domain specialists that degrade significantly outside their training distribution, with F1 drops exceeding 70% across domains. Deploying multiple specialized models is impractical, yet training a single cross-domain detector requires simultaneous access to all training data. To address this, we investigate model merging, a family of post-hoc techniques that operate solely on trained checkpoints. We evaluate parameter merging with five task-vector methods, architecture merging via greedy layer stitching, and cross-tokenizer alignment across four code models, three benchmarks, and twelve configurations. Same-base TIES merging creates effective cross-domain detectors, validated across two model families and three random seeds, reaching 0.865 combined F1 on UniXcoder, 93% of multi-task performance without any training data at the merging step. WUDI achieves the highest in-distribution combined F1 at 0.899, but TIES generalizes better to unseen AI-generated clones, making it our recommended method. Cross-base merging yields only marginal and high-variance gains across all five methods, indicating that task vector compatibility through a shared pre-trained base is the binding factor for effective merging. Merged detectors also outperform zero-shot code LLMs on GPTCloneBench at lower inference cost and generalize up to 4x better than multi-task training to unseen AI-generated clones, suggesting a trade-off between in-domain performance and OOD robustness. This work provides one of the first systematic empirical studies of model merging for software engineering and a practical recipe for building cross-domain clone detectors.

What, Where, and How: Disentangling the Roles of Task, Language, and Model in Code Model Representations

Do independently trained language models come to represent the same thing in the same way? We answer for code, extending a recently introduced concept-circuit extraction method to a 2x2 design -- Python and Rust crossed with Qwen2.5-Coder-7B and DeepSeek-Coder-V1-6.7B -- and measuring a complete inventory of grammatical concepts (58 Python, 57 Rust) identically in all four cells: the smallest design that separates what depends on the task, the language, and the model. The answer splits into three parts. What earns dedicated circuitry is set by the task: the models agree on which concepts receive circuits (Spearman $ρ$ = 0.638 for Python, 0.673 for Rust, both p < $10^{-7}$). Where those circuits sit is set by the model: Qwen processes concepts in a late band (~L17-19), DeepSeek at L6-7, for both languages. How circuits grow across layers is also set by the model: Qwen gives its atomic concepts an early spike that DeepSeek does not. "Are circuits universal?" thus has no single answer: yes for What, no for Where and How -- universality is a property of representational content, not of computational organisation. None of this structure was fixed in advance. The agreement could have landed anywhere between independence and identity; it lands at $ρ\approx 0.65$. Rust constructs receive 2-3x more concept-specific circuitry than their Python equivalents, in both models. Both models share neurons between the languages (6/7 and 7/7 paired constructs), DeepSeek 1.94x more than Qwen -- a direction no prior result predicts. And Qwen binds nine keywords of Rust's type-and-trait machinery into one tight neuron cluster (Jaccard 0.535 vs null 0.112, p < 0.001), a semantic dimension invisible in surface syntax. Ablation and linear probes confirm the circuits are functional. All claims are scoped to this 2x2; whether the per-model profile predicts a third model is the designed next test.

TypeProbe: Recovering Type Representations from Hidden States of Pre-trained Code Models

State-of-the-art code models achieve impressive performance, yet the extent to which they internally encode type information remains poorly understood. We probe the residual streams of pretrained code models for internal type representations using a parallel dataset of Java and Python code examples. Our results show that cross-lingual type representations emerge even from untyped code. Moreover, we test whether hidden states linearly encode the result type implied by typed function application by training probes on one language to infer argument and result types in the other. Finally, we find that this structure is partly robust to lexical perturbations and cross-language syntactic variations. To the best of our knowledge, prior work on interpretability of code models has not directly targeted formal type semantics or cross-lingual type representations. We release our code and datasets.

Falsification, Not Exposure: An Internally Preregistered Placebo-Controlled Decomposition of Self-Repair Feedback in Frozen Small Code Models

In deployment settings where retraining is infeasible, small frozen code models are routinely asked to repair a failed program after seeing their own failing output, usually treated as a retry mechanism. From a Popperian view, a generated program is a conjecture and a test-execution violation is an oracle-relative, executable counterexample, so feedback's value should be attributed not to re-exposure to failing code but to whether the conjecture is opened to external, executable criticism. As the third stage of a falsification-centered measurement program, this study builds a placebo-controlled instrument that decomposes the feedback packet against a blind-resampling baseline at matched output-generation budget and against content-free, shape-matched placebos. The contribution is not a new repair algorithm but a reflexive methodology (packet decomposition, placebo mirroring, matched-budget discordant-pair tests, fresh-generation confirmation, executable audits) that makes both the model's program conjecture and the researcher's "feedback content works" claim falsifiable. Across six HumanEval+/MBPP+ cells with three 0.5B-1.5B frozen models, 290 dead task-cell units (no best-of-8 candidate passing the public tier) were evaluated; the main run produced 7,000 fresh generations and a preregistered follow-up 1,400 more. Blind resampling exceeded bare-code retry by +18 net unlocks (25/7, Holm p=0.0021). Code-plus-facts recovered +18 over bare code (21/3, p=0.00042) and +15 over a generic-bullet placebo (p=0.0041). An instruction-only effect was not distinguishable (+3, p=0.36). Code-plus-facts and blind resampling tied at 26 unlocks each (not equivalence). Six external-controller follow-ups tied a content-free shape placebo. In this regime, falsification helped not as vocabulary or self-critique, but as comparison with external, executable counterexamples.

SamatNext v0.2-B: An Exploratory Study of RMS-Normalized Hybrid Decoders for Curriculum Retention in Small Code Models

Standard autoregressive Transformer decoders can often exhibit substantial forgetting under sequential fine-tuning on shifting curriculum distributions. This technical report evaluates SamatNext v0.2-B, an experimental 356M-parameter hybrid sequence decoder that alternates Differential-Attention-style layers with DeltaNet-inspired simplified linear-state mixer layers using RMS normalization and output scale calibration. We study the model under a controlled staged Python code curriculum and compare it with a parameter-matched Transformer baseline. In this setting, SamatNext v0.2-B achieves a 100.0% pass rate on the controlled Stage 5 holdout while retaining 98.8% of adjacent Stage 3 semantic behavior and reaching 12.0% on the Stage 2E early syntax holdout. The strongest Transformer baseline reaches 97.6% on Stage 5 but retains only 6.0% of Stage 3 behavior. Both architectures remain weak on long-horizon early-stage retention, so the result should be interpreted as evidence of an altered retention/plasticity tradeoff in this controlled setting, not as a general solution to catastrophic forgetting. Code, model specifications, evaluation scripts, and result tables are provided for independent verification.