LLM Post-Training

Post-training data plays a pivotal role in shaping the capabilities of Large Language Models (LLMs), yet datasets are often treated as isolated artifacts, overlooking the systemic connections that underlie their evolution. To disentangle these complex relationships, we introduce the concept of data lineage to the LLM ecosystem and propose an automated multi-agent framework to reconstruct the evolutionary graph of dataset development. Through large-scale lineage analysis, we characterize domain-specific structural patterns, such as vertical refinement in math-oriented datasets and horizontal aggregation in general-domain corpora. Moreover, we uncover pervasive systemic issues, including structural redundancy induced by implicit dataset intersections and the propagation of benchmark contamination along lineage paths. To demonstrate the practical value of lineage analysis for data construction, we leverage the reconstructed lineage graph to create a lineage-aware diversity-oriented dataset. By anchoring instruction sampling at upstream root sources, this approach mitigates downstream homogenization and hidden redundancy, yielding a more diverse post-training corpus. We further highlight lineage-centric analysis as an efficient and robust topological alternative to sample-level dataset comparison for large-scale data ecosystems. By grounding data construction in explicit lineage structures, our work advances post-training data curation toward a more systematic and controllable paradigm.

While Experience Replay - the practice of storing rollouts and reusing them multiple times during training - is a foundational technique in general RL, it remains largely unexplored in LLM post-training due to the prevailing belief that fresh, on-policy data is essential for high performance. In this work, we challenge this assumption. We present a systematic study of replay buffers for LLM post-training, formalizing the optimal design as a trade-off between staleness-induced variance, sample diversity and the high computational cost of generation. We show that strict on-policy sampling is suboptimal when generation is expensive. Empirically, we show that a well-designed replay buffer can drastically reduce inference compute without degrading - and in some cases even improving - final model performance, while preserving policy entropy.

Heavy supervised fine-tuning on a target domain can strongly suppress capabilities that were present in the base model. We study this phenomenon in formal mathematics using Goedel-Prover-V2, an open-source model heavily trained on 1.8 million formal-math examples. After domain specialization, the model almost completely loses its ability to produce valid tool calls, even when explicitly instructed to use tools, dropping from 89.4% function-calling accuracy in the base model to nearly 0%. We ask whether this agentic collapse is permanent or instead reversible. To answer this question, we fine-tune the specialized model on a small amount of Lean-specific tool-use data. Remarkably, as few as 100 agentic traces are sufficient to restore strong tool-calling behavior. Importantly, this recovery is not the result of reward hacking or benchmark-specific optimization: the recovery data is entirely drawn from the Lean setting, where the model uses natural-language queries to search the Mathlib library for relevant theorems and lemmas, yet the regained capability transfers well beyond that domain. In particular, these same 100 Lean-specific traces improve performance on the Berkeley Function Calling Leaderboard from near zero to 83.8%, approaching the base model's 89.4% despite the mismatch in task distribution and protocol. The recovered capability is also practically useful in-domain. On ProofNet, pass@32 improves from 21.51% to 25.81%. Together, these results show that heavy domain supervised fine-tuning can suppress general tool-use ability without permanently erasing it, and that a small amount of domain-specific agentic data can awaken dormant tool-use capabilities.

Post-training has become central to turning pretrained large language models (LLMs) into aligned, capable, and deployable systems. Recent progress spans supervised fine-tuning (SFT), preference optimization, reinforcement learning (RL), process supervision, verifier-guided methods, distillation, and multi-stage pipelines. Yet these methods are often discussed in fragmented ways, organized by labels or objectives rather than by the behavioral bottlenecks they address. This survey argues that LLM post-training is best understood as structured intervention on model behavior. We organize the field first by trajectory provenance, which defines two primary regimes: off-policy learning on externally supplied trajectories and on-policy learning on learner-generated rollouts. We then interpret methods through two recurring roles -- effective support expansion, which makes useful behaviors more reachable, and policy reshaping, which improves behavior within already reachable regions -- together with a complementary systems-level role, behavioral consolidation, which preserves, transfers, and amortizes useful behavior across stages and model transitions. Under this view, SFT may serve either support expansion or policy reshaping; preference optimization is usually off-policy reshaping, though online variants move closer to learner-generated states. On-policy RL often improves behavior on learner-generated states, but stronger guidance can also make hard-to-reach reasoning paths reachable. Distillation is often better understood as consolidation rather than only compression, and hybrid pipelines emerge as coordinated multi-stage compositions. Overall, the framework helps diagnose post-training bottlenecks and reason about stage composition, suggesting that progress increasingly depends on coordinated systems design rather than any single dominant objective.

The deployment of large language models (LLMs) raises significant ethical and safety concerns. While LLM alignment techniques are adopted to improve model safety and trustworthiness, adversaries can exploit these techniques to undermine safety for malicious purposes, resulting in misalignment. Misaligned LLMs may be published on open platforms to magnify harm. To address this, additional safety alignment, referred to as realignment, is necessary before deploying untrusted third-party LLMs. This study explores the efficacy of fine-tuning methods in terms of misalignment, realignment, and the effects of their interplay. By evaluating four Supervised Fine-Tuning (SFT) and two Preference Fine-Tuning (PFT) methods across four popular safety-aligned LLMs, we reveal a mechanism asymmetry between attack and defense. While Odds Ratio Preference Optimization (ORPO) is most effective for misalignment, Direct Preference Optimization (DPO) excels in realignment, albeit at the expense of model utility. Additionally, we identify model-specific resistance, residual effects of multi-round adversarial dynamics, and other noteworthy findings. These findings highlight the need for robust safeguards and customized safety alignment strategies to mitigate potential risks in the deployment of LLMs. Our code is available at https://github.com/zhangrui4041/The-Art-of-Mis-alignment.

Reinforcement learning (RL) has become a central post-training tool for improving the reasoning abilities of large language models (LLMs). In these systems, the rollout, the trajectory sampled from a prompt to termination, including intermediate reasoning steps and optional tool or environment interactions, determines the data the optimizer learns from, yet rollout design is often underreported. This survey provides an optimizer-agnostic view of rollout strategies for RL-based post-training of reasoning LLMs. We formalize rollout pipelines with unified notation and introduce Generate-Filter-Control-Replay (GFCR), a lifecycle taxonomy that decomposes rollout pipelines into four modular stages: Generate proposes candidate trajectories and topologies; Filter constructs intermediate signals via verifiers, judges, critics; Control allocates compute and makes continuation/branching/stopping decisions under budgets; and Replay retains and reuses artifacts across rollouts without weight updates, including self-evolving curricula that autonomously generate new training tasks. We complement GFCR with a criterion taxonomy of reliability, coverage, and cost sensitivity that characterizes rollout trade-offs. Using this framework, we synthesize methods spanning RL with verifiable rewards, process supervision, judge-based gating, guided and tree/segment rollouts, adaptive compute allocation, early-exit and partial rollouts, throughput optimization, and replay/recomposition for self-improvement. We ground the framework with case studies in math, code/SQL, multimodal reasoning, tool-using agents, and agentic skill benchmarks that evaluate skill induction, reuse, and cross-task transfer. Finally, we provide a diagnostic index that maps common rollout pathologies to GFCR modules and mitigation levers, alongside open challenges for building reproducible, compute-efficient, and trustworthy rollout pipelines.

In the field of software operations, Large Language Models (LLMs) have attracted increasing attention. However, existing research has not yet achieved efficient and effective endto-end intelligent operations due to low-quality data, fragmented knowledge and insufficient learning. To explore the potential of LLMs in software operations, we propose OpsLLM, a domainspecific LLM that supports both knowledge-based question answering (QA) and root cause analysis (RCA). Moreover, we disclose the detailed workflow for building LLMs specifically in the software operations domain. First, a Human-in-the-Loop mechanism is introduced to curate high-quality data from a large collection of operational data and construct a fine-tuning dataset. Then, based on the data, supervised fine-tuning is conducted to achieve a base model. Furthermore, we introduce a domain process reward model (DPRM) during the reinforcement learning stage to optimize the accuracy and reliability of the fine-tuned model on RCA tasks. Experimental results on the tasks with diverse difficulties demonstrate that OpsLLMs effectively learns and aligns with the operational domain knowledge infused, outperforming existing open-source and closed-source LLMs in accuracy with improvements of 0.2%~11.9% on QA tasks and 8.5%~70.3% on RCA tasks, while exhibiting strong transferability. Moreover, we will open-source three versions of OpsLLM with 7B, 14B and 32B parameters, along with a 15K fine-tuning dataset.

Anticipating supply chain disruptions before they materialize is a core challenge for firms and policymakers alike. A key difficulty is learning to reason reliably about infrequent, high-impact events from noisy and unstructured inputs - a setting where general-purpose models struggle without task-specific adaptation. We introduce an end-to-end framework that trains LLMs to produce calibrated probabilistic forecasts using realized disruption outcomes as supervision. The resulting model substantially outperforms strong baselines - including GPT-5 - on accuracy, calibration, and precision. We also show that training induces more structured and reliable probabilistic reasoning without explicit prompting. These results suggest a general pathway for training domain-specific forecasting models that produce decision-ready signals. To support transparency we open-source the evaluation dataset used in this study. Dataset: https://huggingface.co/datasets/LightningRodLabs/supply-chain-predictions

As Large Language Models (LLMs) are increasingly adopted in edge intelligence to power domain-specific applications and personalized services, the quality and efficiency of the LLM post-training phase-including fine-tuning and inference, have become critical due to constrained resources. Although recent advances in federated parameter-efficient fine-tuning (FL PEFT) and low-latency inference have improved individual task performance, fine-tuning and inference are still handled as isolated workloads, which overlooks their interdependence and results in redundant deployments and delayed improvement in inference quality. To address these limitations, we introduce a new co-execution framework and instantiate it with CoLLM, a system that unifies FL PEFT and inference on shared edge replicas and model parameters. CoLLM addresses key challenges at both replica and cluster levels through: (1) an intra-replica model sharing mechanism that enables real-time model parameter reuse via unmerged inference and shadow adapter strategies; and (2) a two-timescale inter-replica coordination algorithm that adaptively balances fine-tuning and inference workloads to jointly optimize long-term model quality gains and short-term inference efficiency. Extensive evaluation across diverse LLMs and real-world traces show that CoLLM consistently outperforms state-of-the-art LLM systems, achieving up to 3x higher goodput, demonstrating its effectiveness in enabling seamless LLM post-training for edge intelligence.

Register Transfer Level (RTL) design translates high-level specifications into hardware using HDLs such as Verilog. Although LLM-based RTL generation is promising, the scarcity of functionally verifiable high-quality data limits both accuracy and diversity. Existing post-training typically produces a single HDL implementation per specification, lacking awareness of RTL variations needed for different design goals. We propose RTLSeek, a post-training paradigm that applies rule-based Diversity-Oriented Reinforcement Learning to improve RTL correctness and diversity. Our Diversity-Centric Multi-Objective Reward Scheduling integrates expert knowledge with EDA feedback, and a three-stage framework maximizes the utility of limited data. Experiments on the RTLLM benchmark show that RTLSeek surpasses prior methods, with ablation results confirming that encouraging broader design-space exploration improves RTL quality and achieves the principle of "the more generated, the better results." Implementation framework, including the dataset, source code, and model weights, is shown at https://anonymous.4open.science/r/DAC2026ID71-ACB4/.

The ability to precisely derive mathematical objects is a core requirement for downstream STEM applications, including mathematics, physics, and chemistry, where reasoning must culminate in formally structured expressions. Yet, current LM evaluations of mathematical and scientific reasoning rely heavily on simplified answer formats such as numerical values or multiple choice options due to the convenience of automated assessment. In this paper we provide three contributions for improving reasoning over mathematical objects: (i) we build and release training data and benchmarks for deriving mathematical objects, the Principia suite; (ii) we provide training recipes with strong LLM-judges and verifiers, where we show that on-policy judge training boosts performance; (iii) we show how on-policy training can also be used to scale test-time compute via aggregation. We find that strong LMs such as Qwen3-235B and o3 struggle on Principia, while our training recipes can bring significant improvements over different LLM backbones, while simultaneously improving results on existing numerical and MCQA tasks, demonstrating cross-format generalization of reasoning abilities.

LLM post-training pipelines that combine supervised fine-tuning and reinforcement learning are difficult to configure under realistic compute budgets: the configuration space is high-dimensional and heterogeneous, stages are strongly coupled, and each end-to-end evaluation is expensive. We propose AutoPipe, a budget-aware two-phase framework for configuration selection in LLM post-training. Offline, AutoPipe learns a dataset-conditioned learning-to-rank surrogate from historical runs, capturing within-dataset preferences and providing transferable guidance toward promising regions of the configuration space. Online, for a new dataset, AutoPipe uses the offline guidance to steer Bayesian optimization and models dataset-specific deviations with a Gaussian-process residual surrogate. To reduce evaluation cost, each trial is early-stopped and scored by a learned predictor that maps early training signals to a low-cost proxy for final post-training performance. Experiments on biomedical reasoning tasks show that AutoPipe achieves performance comparable to full-fidelity HPO baselines at a fraction of their online cost, while outperforming offline-only approaches and early-stopped HPO variants.

LLM agents are becoming increasingly important in the security domain, but leading systems are often closed-source, cloud-based, hard to reproduce or use with sensitive code. This creates a need for small, local models that can perform security tasks under strict resource constraints, though effective methods for developing them remain unexplored. In this paper, we address this gap by proposing a two-stage post-training recipe that turns a small local language model into a security agent. To this end, we focus on Linux privilege escalation as a representative setting to systematically study the training of local models, as the task is both automatically verifiable and requires multi-step interactive reasoning. Using an experimental setup that mitigates data leakage, we post-train a small 4B model in two stages: supervised fine-tuning on traces from procedural privilege-escalation environments, followed by reinforcement learning with verifiable rewards. On a held-out benchmark of 12 Linux privilege-escalation scenarios, supervised fine-tuning doubles the baseline success rate under a tight budget of 20 interaction rounds, and subsequent reinforcement learning training improves our model, PrivEsc-LLM 4B, to 93.3% success, behind only Claude Opus 4.7 at this budget. At the same time, the expected inference cost per successful escalation decreases by more than 80x. Our findings not only show that small local models can be adapted to complex security tasks, but also document the challenges involved, offering guidance for transferring this recipe to other settings.

Technical-service LLM agents are entering production workflows, where value depends on whether engineers adopt generated replies. Service tickets hide decision logic, contain noisy single-reference responses, and make reward evaluation costly, making standard post-training brittle. Existing post-training and LLM-as-a-Judge approaches improve grounding or feedback, but do not jointly model latent decision logic, response diversity, and reward cost. We address this gap by coupling latent logic augmentation, robust noise reduction, and hybrid reward modeling. The framework augments supervised fine-tuning data with Planning-Aware Trajectory Modeling and Reasoning Augmentation, builds dual-filtered Multiple Ground Truths, and trains the policy with a hybrid reward that combines a Reranker with an LLM-as-a-Judge. On real Cloud technical-service tasks, the adapted Qwen3-4B achieves the highest Multi-ECS (0.441), lower reward cost, and the highest production adoption rate (46.63%).

Pre-trained Large Language Model (LLM) exhibits broad capabilities, yet, for specific tasks or domains their attainment of higher accuracy and more reliable reasoning generally depends on post-training through Supervised Fine-Tuning (SFT) or Reinforcement Learning (RL). Although often treated as distinct methodologies, recent theoretical and empirical developments demonstrate that SFT and RL are closely connected. This study presents a comprehensive and unified perspective on LLM post-training with SFT and RL. We first provide an in-depth overview of both techniques, examining their objectives, algorithmic structures, and data requirements. We then systematically analyze their interplay, highlighting frameworks that integrate SFT and RL, hybrid training pipelines, and methods that leverage their complementary strengths. Drawing on a representative set of recent application studies from 2023 to 2025, we identify emerging trends, characterize the rapid shift toward hybrid post-training paradigms, and distill key takeaways that clarify when and why each method is most effective. By synthesizing theoretical insights, practical methodologies, and empirical evidence, this study establishes a coherent understanding of SFT and RL within a unified framework and outlines promising directions for future research in scalable, efficient, and generalizable LLM post-training.

Pretraining produces a learned parameter vector that is typically treated as a starting point for further iterative adaptation. In this work, we instead view the outcome of pretraining as a distribution over parameter vectors, whose support already contains task-specific experts. We show that in small models such expert solutions occupy a negligible fraction of the volume of this distribution, making their discovery reliant on structured optimization methods such as gradient descent. In contrast, in large, well-pretrained models the density of task-experts increases dramatically, so that diverse, task-improving specialists populate a substantial fraction of the neighborhood around the pretrained weights. Motivated by this perspective, we explore a simple, fully parallel post-training method that samples $N$ parameter perturbations at random, selects the top $K$, and ensembles predictions via majority vote. Despite its simplicity, this approach is competitive with standard post-training methods such as PPO, GRPO, and ES for contemporary large-scale models.

Continual post-training of generative models is widely used, yet a principled understanding of when and why forgetting occurs remains limited. We develop theoretical results under a two-mode mixture abstraction (representing old and new tasks), proposed by Chen et al. (2025) (arXiv:2510.18874), and formalize forgetting in two forms: (i) mass forgetting, where the old mixture weight collapses to zero, and (ii) old-component drift, where an already-correct old component shifts during training. For equal-covariance Gaussian modes, we prove that forward-KL objectives trained on data from the new distribution drive the old weight to zero, while reverse-KL objectives converge to the true target (thereby avoiding mass forgetting) and perturb the old mean only through overlap-gated misassignment probabilities controlled by the Bhattacharyya coefficient, yielding drift that decays exponentially with mode separation and a locally well-conditioned geometry with exponential convergence. We further quantify how replay interacts with these objectives. For forward-KL, replay must modify the training distribution to change the population optimum; for reverse-KL, replay leaves the population objective unchanged but prevents finite-batch old-mode starvation through bounded importance weighting. Finally, we analyze three recently proposed near-on-policy post-training methods, SDFT (arxiv:2601.19897), TTT-Discover (arxiv:2601.16175), and OAPL (arxiv:2602.19362), via the same lens and derive explicit conditions under which each retains old mass and exhibits overlap-controlled drift. Overall, our results show that forgetting can by precisely quantified based on the interaction between divergence direction, geometric behavioral overlap, sampling regime, and the visibility of past behavior during training.

While post-training has successfully improved large language models (LLMs) across a variety of domains, these gains heavily rely on human-labeled data or external verifiers. Existing data has already been exploited, and new data is expensive to collect. Moreover, true intelligence goes far beyond verifiable tasks. Therefore, we need self-improvement frameworks that are less dependent on external signals and more broadly applicable to both verifiable and non-verifiable domains. We propose Mutual Information Preference Optimization (MIPO), a contrastive data augmentation method that constructs preference pairs by generating a positive response conditioning on the correct prompt, and a negative response by conditioning on a random, unrelated prompt. We show that using Direct Preference Optimization to learn from this paired data maximizes pointwise mutual information under the base LLM between prompts and model responses. Experiments with with 1-7B parameter Llama and Qwen instruct models show that MIPO achieves 3-16% gains (and 51% increase for Qwen2.5-1.5B-Instruct) on personalization compared to prompting baselines. Surprisingly, MIPO can also be useful in verifiable domains, such as math and multiple-choice question answering, yielding 1-20% gains without any additional data or external supervision. These results suggest a promising direction for self-improvement using intrinsic signals derived from contrastive data pairs.

Despite the impressive performance of general-purpose large language models (LLMs), they often require fine-tuning or post-training to excel at specific tasks. For instance, large reasoning models (LRMs), such as the DeepSeek-R1 series, demonstrate strong reasoning capabilities after post-training different general large language models on diverse chain-of-thought (CoT) datasets. However, this additional training frequently comes at the cost of reduced safety, as the fine-tuned or post-trained models tend to exhibit more harmful behaviors compared with the regular LLMs before post-training or fine-tuning, potentially leading to harmful outcomes due to their enhanced capabilities. Taking LRMs as an example, we first investigate the underlying cause of this safety degradation in this paper. Our analysis reveals that post-training can mask the original safety mechanisms of the base LLM, while over-amplifying representations related to their post-training ability. But luckily, we also find that LRMs' safety mechanisms still exist instead of being removed during their post-training. Based on these findings, we propose a lightweight and cost-effective solution called SafeReAct that restores the suppressed safety behaviors by aligning with LoRA adapters on a few layers. Experiments on four state-of-the-art LRMs show that our method significantly improves safety on harmful prompts without compromising reasoning performance. Besides LRMs, additional results on other domain-specific LLMs, like medical models, further confirm the generality and effectiveness of our approach.

AI agents have become surprisingly proficient at software engineering over the past year, largely due to improvements in reasoning capabilities. This raises a deeper question: can these systems extend their capabilities to automate AI research itself? In this paper, we explore post-training, the critical phase that turns base LLMs into useful assistants. We introduce PostTrainBench to benchmark how well LLM agents can perform post-training autonomously under bounded compute constraints (10 hours on one H100 GPU). We ask frontier agents (e.g., Claude Code with Opus 4.6) to optimize the performance of a base LLM on a particular benchmark (e.g., Qwen3-4B on AIME). Importantly, we do not provide any predefined strategies to the agents and instead give them full autonomy to find necessary information on the web, run experiments, and curate data. We find that frontier agents make substantial progress but generally lag behind instruction-tuned LLMs from leading providers: 23.2% for the best agent vs. 51.1% for official instruction-tuned models. However, agents can exceed instruction-tuned models in targeted scenarios: GPT-5.1 Codex Max achieves 89% on BFCL with Gemma-3-4B vs. 67% for the official model. We also observe several failure modes worth flagging. Agents sometimes engage in reward hacking: training on the test set, downloading existing instruction-tuned checkpoints instead of training their own, and using API keys they find to generate synthetic data without authorization. These behaviors are concerning and highlight the importance of careful sandboxing as these systems become more capable. Overall, we hope PostTrainBench will be useful for tracking progress in AI R&D automation and for studying the risks that come with it. Website and code are available at https://posttrainbench.com/.

Large language models (LLMs) are predominantly trained on English-centric data, resulting in uneven performance for smaller languages. We study whether continued pretraining (CPT) can improve Estonian capabilities in multilingual LLMs while preserving English and general reasoning performance. Using Llama 3.1 8B and Apertus 8B as base models, we apply CPT with Estonian-enriched multilingual replay, followed by mostly English supervised fine-tuning, preference optimization, and chat vector merging. Evaluation on Estonian benchmarks, targeted pairwise human evaluation, and an Estonian Chatbot Arena-style setup shows consistent improvements in Estonian language competence, reasoning, translation, and instruction-following. Although Apertus exhibits stronger Estonian capabilities before adaptation, the more English-centric Llama model achieves substantially larger gains after adaptation. While some English capabilities regress relative to the original instruction-tuned models, chat vector merging substantially restores English instruction-following and reasoning performance. These findings suggest that CPT with balanced multilingual replay and lightweight post-training alignment can substantially improve single-language capabilities in multilingual LLMs.

This report presents CharacterFlywheel, an iterative flywheel process for improving large language models (LLMs) in production social chat applications across Instagram, WhatsApp, and Messenger. Starting from LLaMA 3.1, we refined models across 15 generations using data from both internal and external real-user traffic. Through continuous deployments from July 2024 to April 2025, we conducted controlled 7-day A/B tests showing consistent engagement improvements: 7 of 8 newly deployed models demonstrated positive lift over the baseline, with the strongest performers achieving up to 8.8% improvement in engagement breadth and 19.4% in engagement depth. We also observed substantial gains in steerability, with instruction following increasing from 59.2% to 84.8% and instruction violations decreasing from 26.6% to 5.8%. We detail the CharacterFlywheel process which integrates data curation, reward modeling to estimate and interpolate the landscape of engagement metrics, supervised fine-tuning (SFT), reinforcement learning (RL), and both offline and online evaluation to ensure reliable progress at each optimization step. We also discuss our methods for overfitting prevention and navigating production dynamics at scale. These contributions advance the scientific rigor and understanding of LLMs in social applications serving millions of users.

Injecting new reasoning knowledge into Large Language Models (LLMs) via post-training often induces catastrophic forgetting. Recent studies emphasize the importance of on-policy data but suggest that KL-divergence fails to mitigate forgetting. In contrast, we show, both analytically and empirically, that the KL-constrained reward formulation actually plays a critical role in retaining knowledge during post-training. This motivates our Surgical Post-Training (SPOT), a proximal on-policy distillation framework designed to optimize reasoning efficiently while preserving prior knowledge. SPOT consists of (1) a data rectification pipeline employing an Oracle to surgically correct erroneous steps via minimal edits, generating proximal on-policy data; and (2) a reward-based binary cross-entropy objective essential for enhancing reasoning and mitigating forgetting. Empirically, with only 4k rectified math pairs, SPOT improves Qwen3-8B's accuracy by 6.2% on average across in-domain and out-of-domain tasks, requiring merely 16-minute model training on 8x H800 GPUs. Moreover, SPOT provides a superior initialization for subsequent reinforcement learning, significantly elevating the performance ceiling. Code: https://github.com/Visual-AI/SPoT

Knowledge graphs provide structured and reliable information for many real-world applications, motivating increasing interest in combining large language models (LLMs) with graph-based retrieval to improve factual grounding. Recent Graph-based Retrieval-Augmented Generation (GraphRAG) methods therefore introduce iterative interaction between LLMs and knowledge graphs to enhance reasoning capability. However, existing approaches typically depend on manually designed guidance and interact with knowledge graphs through a limited set of predefined tools, which substantially constrains graph exploration. To address these limitations, we propose GraphScout, a training-centric agentic graph reasoning framework equipped with more flexible graph exploration tools. GraphScout enables models to autonomously interact with knowledge graphs to synthesize structured training data which are then used to post-train LLMs, thereby internalizing agentic graph reasoning ability without laborious manual annotation or task curation. Extensive experiments across five knowledge-graph domains show that a small model (e.g., Qwen3-4B) augmented with GraphScout outperforms baseline methods built on leading LLMs (e.g., Qwen-Max) by an average of 16.7% while requiring significantly fewer inference tokens. Moreover, GraphScout exhibits robust cross-domain transfer performance. Our code will be made publicly available~\footnote{https://github.com/Ying-Yuchen/_GraphScout_}.

The transition of Large Language Models (LLMs) from exploratory tools to active "silicon subjects" in social science lacks extensive validation of operational validity. This study introduces Conditioned Comment Prediction (CCP), a task in which a model predicts how a user would comment on a given stimulus by comparing generated outputs with authentic digital traces. This framework enables a rigorous evaluation of current LLM capabilities with respect to the simulation of social media user behavior. We evaluated open-weight 8B models (Llama3.1, Qwen3, Ministral) in English, German, and Luxembourgish language scenarios. By systematically comparing prompting strategies (explicit vs. implicit) and the impact of Supervised Fine-Tuning (SFT), we identify a critical form vs. content decoupling in low-resource settings: while SFT aligns the surface structure of the text output (length and syntax), it degrades semantic grounding. Furthermore, we demonstrate that explicit conditioning (generated biographies) becomes redundant under fine-tuning, as models successfully perform latent inference directly from behavioral histories. Our findings challenge current "naive prompting" paradigms and offer operational guidelines prioritizing authentic behavioral traces over descriptive personas for high-fidelity simulation.

Table Question Answering (TQA) aims to answer natural language questions over structured tables. Large Language Models (LLMs) enable promising solutions to this problem, with operator-centric solutions that generate table manipulation pipelines in a multi-step manner offering state-of-the-art performance. However, these solutions rely on multiple LLM calls, resulting in prohibitive latencies and computational costs. We propose Operation-R1, the first framework that trains lightweight LLMs (e.g., Qwen-4B/1.7B) via a novel variant of reinforcement learning with verifiable rewards to produce high-quality data-preparation pipelines for TQA in a single inference step. To train such an LLM, we first introduce a self-supervised rewarding mechanism to automatically obtain fine-grained pipeline-wise supervision signals for LLM training. We also propose variance-aware group resampling to mitigate training instability. To further enhance robustness of pipeline generation, we develop two complementary mechanisms: operation merge, which filters spurious operations through multi-candidate consensus, and adaptive rollback, which offers runtime protection against information loss in data transformation. Experiments on two benchmark datasets show that, with the same LLM backbone, Operation-R1 achieves average absolute accuracy gains of 8.83 and 4.44 percentage points over multi-step preparation baselines, with 79% table compression and a 2.2$\times$ reduction in monetary cost.

Large language model (LLM) post-training enhances latent skills, unlocks value alignment, improves performance, and enables domain adaptation. Unfortunately, post-training is known to induce forgetting, especially in the ubiquitous use-case of leveraging third-party pre-trained models, which is typically understood as a loss of parametric or factual knowledge. We argue that this accuracy-centric view is insufficient for modern foundation models and instead define forgetting as systematic model drift that degrades behavior and user experience. In this context, we introduce CapTrack, a capability-centric framework for analyzing forgetting in LLMs that combines a behavioral taxonomy with an evaluation suite centered on capability-specific metrics. Using CapTrack, we conduct a large-scale empirical study across post-training algorithms, domains, and model families, including models up to 80B parameters. We find that forgetting extends beyond parametric knowledge, with pronounced drift in robustness and default behaviors. Instruction fine-tuning induces the strongest relative drift, while preference optimization is more conservative and can partially recover lost capabilities. Differences across model families persist, and no universal mitigation emerges.

Software testing is crucial for ensuring the correctness and reliability of software systems. Automated generation of issue reproduction tests from natural language issue descriptions enhances developer productivity by simplifying root cause analysis, promotes test-driven development -- "test first, write code later", and can be used for improving the effectiveness of automated issue resolution systems like coding agents. Existing methods proposed for this task predominantly rely on closed-source LLMs, with limited exploration of open models. To address this, we propose SWE-Tester -- a novel pipeline for training open-source LLMs to generate issue reproduction tests. First, we curate a high-quality training dataset of 41K instances from 2.6K open-source GitHub repositories and use it to train LLMs of varying sizes and families. The fine-tuned models achieve absolute improvements of up to 10% in success rate and 21% in change coverage on SWT-Bench Verified. Further analysis shows consistent improvements with increased inference-time compute, more data, and larger models. These results highlight the effectiveness of our framework for advancing open-source LLMs in this domain.

Unified multimodal generation architectures that jointly produce text and images have recently emerged as a promising direction for text-to-image (T2I) synthesis. However, many existing systems rely on explicit modality switching, generating reasoning text before switching manually to image generation. This separate, sequential inference process limits cross-modal coupling and prohibits automatic multimodal generation. This work explores post-training to achieve fully unified text-image generation, where a model autonomously transitions from textual reasoning to visual synthesis within a single inference process. We study this on BAGEL, a 14B mixture-of-transformers model that pairs autoregressive text generation with flow-matching image synthesis. We examine the impact of joint text-image generation on T2I performance and the relative importance of each modality during post-training. We additionally explore different post-training data strategies, showing that a targeted dataset addressing specific limitations achieves superior results compared to broad image-caption corpora or benchmark-aligned data. Using offline, reward-weighted post-training with fully self-generated synthetic data, our approach enables improvements in multimodal image generation across four diverse, independent T2I benchmarks, demonstrating the effectiveness of reward-weighting both modalities and strategically designed post-training data.

Engineering LLMs to accelerate life sciences research requires a robust alignment with biomedical knowledge. We observe that biomedical text exhibits a fundamentally different uncertainty structure from general text: dense low-confidence runs encode epistemic knowledge gaps (dense causal chains, rare entities) rather than the sparse aleatoric stylistic variation typical of general text. Based on this discovery, we propose Balanced Fine-Tuning (BFT), a dual-scale post-training method that combines group-normalized token reweighting with sequence-level reallocation toward knowledge-dense samples exhibiting dense epistemic uncertainty. Across medical evaluation, biological reasoning, sparse-reward RL, and biological representation tasks, BFT provides more consistent gains than SFT and DFT under a shared training setup. When replacing the default closed-source backbones in GeneAgent (GPT-4o) and VCWorld (Gemini-2.5-Flash), the BFT-aligned 70B model delivers stronger performance across biological process reasoning and chemical perturbation prediction. Critically, all BFT variants further improve after subsequent GRPO with sparse rewards, while SFT and DFT degrade, suggesting that epistemic-aware post-training provides a more robust policy initialization. Beyond text generation, BFT-aligned LLMs produce more accurate and professional biomedical profile texts; after encoding these profiles with a text embedding model, the resulting representations support gene-level, cell-level, and perturbation-response tasks, suggesting that BFT-enhanced generation can facilitate biological representation and, in turn, broader biomedical downstream tasks.