LLM Alignment

LLM: Large Language Model

Follow this topic Ask about this topic

Momentum

48 papers in the last four weeks, up 118% on the four weeks before. 0.5% of all new papers.

Jul 13Week of Sep 28

Latest papers 422

Oct 8, 2026cs.CL

Predicting Alignment Generalization with Value Representations

LLM developers post-train their models to exhibit prosocial values and behavioral traits, which are enumerated in an alignment target. However, while recent post-training developments have yielded models that score highly on alignment evaluations, training models on sets of narrow behaviors still influences their behavior across unseen contexts and environments in unexpected ways. In this paper, we establish the task of alignment generalization prediction, i.e., predicting how fine-tuning a model to follow a given value changes its behavior across a wide range of held-out values. We conduct a large-scale analysis of alignment generalization effects across 66 values found in modern alignment targets, and benchmark representational techniques on the alignment generalization prediction task. We find that representations based on model activations when applying values in context significantly outperform methods based on textual descriptions of the values. Specifically, the best activations-based methods achieve correlations of 0.45 with our generalization matrix, compared with 0.05 from description-based baselines. We then show the applicability of representations that predict alignment generalization toward downstream tasks by using them to measure how similar the values in a multi-value alignment target are, which we find is significantly correlated with model robustness. Finally, we show initial evidence towards a shared, model-independent value space, which we use to develop the first taxonomy of LLM values grounded in empirical generalization dynamics. Our work demonstrates the importance of studying value generalization in LLMs and its application toward the more empirical design and training of model behavior.

Andy Liu, Mehar Bhatia, Karolina Stanczak +3

Carnegie Mellon University · Mila - Quebec AI Institute · McGill University +4

LLM Alignment Representation Geometry in Language Models

Oct 7, 2026eess.AS

Conversational Voice Aesthetic Model with Reinforcement Learning from Human Listeners

We introduce Conversational Voice Aesthetic Model, a speech large language model for describing the voice aesthetics of real or synthetic speech responses in natural conversational contexts. Given a context and a response speech, CVAM describes salient moments that characterize the voice and predicts nine categorical attributes spanning gender, pitch, pacing, emotion, and delivery. The key challenge lies in perceptual fields such as emotion and delivery, which are inherently subjective and lack definitive ground truth. Therefore, we collect ~10 human annotations for each of 3k real and synthetic responses derived from the CANDOR corpus. CVAM is supervised finetuned on synthesized aesthetic descriptions and labels, then optimized with Group Relative Policy Optimization on human judgments. Experiments show that CVAM better agrees with human listeners than Gemini 3.1 Pro and open-source speech LLMs, and outperforms single-human-vs.-rest agreement. Together, we demonstrate the importance of grounding voice aesthetics in human perception and propose a principled framework for human alignment.

Xilin Jiang, Shun Zhang, Tejas Jayashankar +2

Columbia University New York, NY, USA · Meta Superintelligence Labs Menlo Park, CA, USA

Speech Language Models LLM Alignment

Oct 7, 2026cs.AI

SafeEvo: Deciphering the Safety Alignment Mechanism and Evolution in Language Models

Safety interpretability advances the study of Large Language Model (LLM) alignment from behavioral constraints driven by data or algorithms towards a deeper understanding of internal mechanisms. However, existing works have focused primarily on safety-related representations, attention heads, or neurons after alignment, while largely overlooking the safety mechanisms in pretrained-only models and their evolution across alignment checkpoints. To address this, we propose SafeEvo, an interpretability framework from the circuit (sparse subgraphs of an LLM) perspective. SafeEvo first applies an optimization-based extraction algorithm to identify weak refusal circuits in pretrained base LLMs that can independently express refusal behavior. Causally ablating these circuits completely eliminates the base model's refusal of harmful inputs. SafeEvo then traces the evolution of refusal circuits across successive alignment checkpoints and finds that their structures change progressively, suggesting that the alignment tax may result from refusal-circuit updates affecting utility-related parameters. To validate this, SafeEvo introduces Safety Circuit Alignment (SCA), which confines safety updates to the refusal circuits. Experiments across three LLMs and two alignment algorithms show that, on average, SCA outperforms vanilla alignment in three aspects: \textbf{(1) stronger alignment}, lowering harmfulness score by 63.21%; \textbf{(2) less over-refusal}, yielding a 58.44% decrease in refusal rates for benign queries; and \textbf{(3) better utility}, retaining 99.58% of the original model capabilities.

Miao Yu, Hao Huang, Lu Yuan +3

The University of Hong Kong (HKU) · Chinese Academy of Sciences (CAS) · Information Engineering University (IEU) +1

LLM Safety Alignment LLM Alignment LLM Interpretability Circuits in Language Models

Oct 7, 2026cs.CL

The Persona Hierarchy Model: Understanding Contextual Generalization in Fine-Tuning LLMs

Language models are routinely fine-tuned under a fixed context, such as a generic system prompt, persona or domain-specific instruction, yet the learned behavior sometimes stays confined to that context and sometimes broadly generalizes to unseen contexts. We propose the Persona Hierarchy Model to explain this: a shared default persona influences behavior across contexts. Under this model, fine-tuning that modifies the shared persona promotes broader transfer, whereas changes to local personas remain more context-specific. Across 120 fine-tuned models spanning four behaviors and 15 training contexts, generalization narrowness positively correlates with the similarity between the training context's persona and the default persona (Pearson's r = 0.72 for Qwen3-4B). Prior fine-tuning under the default context can broaden generalization in subsequent training under other contexts. Aligning contextual responses with default-persona responses produces stronger effects. Finally, we propose persona-preserving regularization (PPR) to confine undesired contextual generalization. In RL, PPR cuts reward hacking from 42-55% to at most 0.2% under every evaluated prompt while retaining accuracy gains. These results support the Persona Hierarchy Model as an explanation for contextual generalization and can motivate future controls on unintended generalization for better alignment of LLMs.

Jiachen Zhao, Zhengxuan Wu, David Bau +1

Northeastern University · Stanford University

LLM Fine-Tuning LLM Alignment Personality Modeling in Language Models

Oct 6, 2026cs.AI

Toward Alignment Scaling Laws: A Framework and First Preregistered Measurements

Whether alignment gets easier or harder as models grow is often argued from isolated findings, as if alignment were one property. We treat it as a family of measurable scaling relations: for each risk category r, the alignment burden needed to hold a fixed safety target is modeled as B_r(N)=a_rN^alpha_r, with N a capability proxy; against a budget proportional to N, scaling helps if alpha_r<1, keeps pace if alpha_r~1, and accumulates alignment debt if alpha_r>1. We give three operationalizations of burden and distinguish observed, audited and true alignment. A toy model, in which corrections consume capability headroom, makes the consequences explicit. We prove that the largest exponent among corrected risks, not an average, sets the long-run regime; that above 1 any policy holding headroom above a floor must grow super-exponentially; that, for burdens that are positive mixtures of power laws, fits on small models underestimate large-scale exponents; and that an audit that uncovers hidden failures without false positives never underestimates true alignment. We propose a pre-registrable protocol and apply reduced versions of it twice. A preregistered reanalysis of public adversarial-training data for Pythia classifiers finds that the compute needed to bring attack success under 10% grows as N^0.60. A preregistered pilot on Qwen2.5 0.5B-72B finds exponents of -0.05 for truthfulness and 0.48 for stated dispositions (both scaling helps under its reduced rule, though local slopes approach 1 at the top; replicated on Qwen3 0.6B-14B), while sycophancy (0.89, or 0.83 with two seeds added at 72B) and a planted backdoor are undetermined: the backdoor is removed quickly when its trigger is known but survives blind safety training at four of five sizes. We release four browser games that play these laws (www.aisafety.fun). We make no claim about which regime holds for current frontier models.

Jeremy Canale

LLM Alignment Language Model Safety Evaluation Language Model Scaling Laws

Oct 5, 2026cs.CL

Reward Stealing Attack on Large Language Models

Adversarial attacks on Large Language Models (LLMs) aim to induce harmful content. However, existing methods suffer from high computational costs or strict model-pairing dependencies, limiting their scalability and transferability. We propose Reward Stealing Attack (ReSA), an adversarial attack framework that targets the latent safety reward underlying LLM alignment. ReSA employs maximum entropy inverse reinforcement learning to recover a proxy reward model solely from the aligned model's behavior. The extracted reward is then reversed at inference time to derive an adversarial policy, efficiently implemented via a reward-guided decoding mechanism. Experiments demonstrate that a single recovered reward generalizes across prompts and diverse models to reveal a fundamental alignment vulnerability, enabling ReSA to significantly outperform existing attacks in effectiveness and transferability. The code is available at https://github.com/GarminQ/ReSA.

Jiaming Qian, Pengyang Zhou, Jiahe Xu +1

Zhejiang University

LLM Alignment Model Extraction Attacks Adversarial Attacks on LLMs Transfer-Based Adversarial Attacks

Oct 4, 2026cs.AI

EmoRSS: Mitigating Emotion-Induced Over-Refusal in Large Language Models

Emotional expression can influence the safety decisions of large language models (LLMs), offering a potential avenue for improving safety alignment. Existing studies have mainly focused on how emotional expressions facilitate attacks under harmful requests, while overlooking their effects on benign requests. We find that emotional expression can also systematically increase refusal tendencies on benign requests, leading to unnecessary over-refusal. Based on this observation, we propose emotion-guided refusal subspace steering (EmoRSS), an activation-steering method that mitigates emotion-induced over-refusal while preserving refusal behaviour on harmful requests. Specifically, we first identify a refusal-sensitive layer using layer-wise linear probes and construct a refusal subspace from sparse autoencoder (SAE) features aligned with the probe direction. Next, we use paired regular and emotional requests with the same queries to estimate the mean activation shift in the features defining the refusal subspace. Finally, we decode this shift into an activation intervention vector and apply it in the reverse refusal direction during inference, without updating the backbone parameters. Experiments on two LLMs show that, when requests contain emotional expressions, our method achieves a more favourable trade-off between refusing harmful requests and answering benign ones than prior over-refusal mitigation baselines, while better preserving general task performance.

Shuyi Miao, Yaojin Ma, Chenhang Cui +5

Beihang University · University of Science and Technology Beijing · National University of Singapore +2

LLM Alignment LLM Refusal Behavior Over-Refusal Activation Steering

Oct 1, 2026cs.SE

Detecting Inconsistencies in Model Specifications with LLM-as-Verifier Reasoning

Model specifications define how large language models (LLMs) should behave, guiding alignment training, inference-time behavior, and evaluation. Yet these specifications may themselves contain defects: two individually reasonable principles may prescribe incompatible behavior when applied to the same situation, leaving no response that satisfies both. Detecting such inconsistencies is challenging. Formalizing natural-language specifications risks losing subtle distinctions, while behavior-based testing cannot reliably distinguish specification defects from differences in model behavior. We introduce VeriSpec, the first approach to directly detect inconsistencies in model specifications by auditing the specification text itself. Our key insight is to preserve the specification in natural language while using an LLM as a verifier. VeriSpec extracts structured, context-aware rules, constructs a topic-guided graph to cluster behaviorally related rules at the same authority level, and applies LLM-as-verifier reasoning to detect inconsistencies. Applying VeriSpec to the OpenAI Model Spec, we extract 405 rules and manually validate five inconsistencies, all reported to its developers, who responded positively and have initiated internal discussions. Compared with five baselines, VeriSpec identifies the most validated inconsistencies, achieves the highest precision (38.5%), and incurs the lowest cost per validated inconsistency ($11.12). These results establish direct specification auditing as a practical complement to behavioral alignment evaluation, catching defects at the source before they shape any model. The code is available at https://github.com/HIPREL-Group/VeriSpec.

Zichen Xie, Mrigank Pawagi, Lize Shao +2

University of Virginia · University of Pennsylvania · The University of Texas at Austin

LLM Alignment LLM Auditing

Oct 1, 2026cs.CL

The Asymptotics of Language Model Alignment with Memory

Language model (LM) alignment broadly aims to perturb a given LM

Q

into an aligned LM

q

such that i) the outputs produced by

q

and

Q

are 'close' in probability, ii)

q

has a higher expected reward than

Q

. Two common techniques for LM alignment are: KL-constrained RL, which requires knowledge of the LM distribution and is computationally expensive, and the best-of-

n

algorithm, which requires only sampling from the LM. The work of Yang et al. established asymptotic closeness between the distributions produced by the two alignment methods for an

m

--length i.i.d. token sequence output by the LM, in the limit as

m

increases to infinity. However, the i.i.d. assumption is not representative of practical LMs, whose output sequences often have memory. In this paper, we extend the asymptotic closeness result to the case when the

m

--length token sequence outputted by the LM is Markovian. Further, for finite-length output sequences -- particularly, when

m=1

-- we provide a complete characterization of LM distributions and reward functions for which the KL-divergence between the distributions produced by the two alignment methods is zero -- a question first posed in Yang et al.

Haricharan Balasundaram, V. Arvind Rameshwar

School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA · IIT Madras · Department of Electrical Engineering, IIT Madras, Chennai, India

RL for Language Models LLM Alignment Language Modeling Inference-Time Language Model Alignment

Oct 1, 2026cs.LG

Beyond Linear Concepts: Discovering and Aligning Non-Linear Concept Manifolds in Large Language Models

Understanding information processing in large language models (LLMs) requires dissecting the geometric organization of their internal token representations. While existing mechanistic interpretability (MI) methods seek to extract concepts, they are constrained by a strong linearity assumption challenged by evidence of non-linear feature manifolds. We move beyond linear concepts by adapting Non-Linear Multi-Dimensional Concept Discovery (NLMCD) from computer vision to token-level LLM activations, modeling concepts as low-dimensional manifolds. To compare concept manifolds across layers and models, we introduce a concept-based alignment (CBA) score, a generalized Rand index that measures geometric proximity without explicit feature matching. Our analysis yields six key findings: (i) a neighboring-layer sanity check shows CBA is more sensitive than PCA- or CKA-based linear baselines; (ii) layer-by-layer alignment matrices reveal two block structures in intermediate and late layers, consistent across models and obscured by linear metrics; (iii) concept composition remains syntax-dominated through most of the network before giving way to increasingly mixed syntactic-semantic concepts in later layers, with increasing output-orientation toward the final layers; (iv) multilingual concept sharing between English and Mandarin is training-dependent rather than universal, strongest in Qwen, weaker in Llama, and absent in GPT-2; (v) inter-model alignment mirrors this structure, with strong correspondence between same-family Qwen models of different scale but weak alignment across model families; and (vi) across Tulu-3 training stages, alignment is highest between adjacent stages, with the largest shift between the base model and SFT, while subsequent preference-alignment stages (DPO, RLVR) leave early layers largely unchanged and RLVR mostly preserves DPO's concepts in late layers.

Tido Specht, Elias Benedict Krey, Nils Neukirch +1

University of Groningen · Carl von Ossietzky University of Oldenburg

LLM Alignment LLM Interpretability Cross-Lingual Representation Alignment Representation Geometry in Language Models

Oct 1, 2026cs.CL

GAW-PO: Preference Optimization with Gradient-Aligned Token Weights

Most preference optimization methods, such as Direct Preference Optimization (DPO), apply preference supervision at the response level, although autoregressive language models are optimized token by token. As a result, all tokens in a rejected response contribute to the negative training signal, including tokens that may encode behavior that is useful for the preferred response. We introduce GAW-PO, a gradient-aligned token reweighting method for DPO that estimates, for each rejected token, whether penalizing it would interfere with the preferred update directions. Tokens whose gradients are strongly aligned with the preferred behavior receive a weaker negative contribution, while conflicting tokens retain a stronger penalty. Our method achieves the highest average performance among the evaluated preference-optimization methods, improving by 0.97 points over standard DPO and 0.65 points over the strongest competing baseline across 11 benchmarks spanning mathematics, reasoning, coding, and question answering. We further show that gradient-aligned weighting is substantially more robust to aggressive preference optimization: as the DPO regularization parameter

β

decreases, standard DPO degrades sharply, whereas GAW-PO continues to improve. These results suggest that accounting for the interaction between rejected-token updates and preferred behavior provides an effective form of token-level credit assignment for preference optimization.

Andreea Dutulescu, Stefan Ruseti, Mihai Masala +2

National University of Science and Technology POLITEHNICA Bucharest · NVIDIA

LLM Alignment Direct Preference Optimization Adaptive Loss Weighting Preference Optimization Preference-Based RL

Sep 30, 2026cs.LG

Pre-training interventions, ex post facto: Grafting model beliefs across checkpoints

Pre-training interventions are critical to alignment research, since beliefs formed during pre-training shape how a model generalizes from later training. One recently popular technique for such interventions is synthetic document fine-tuning (SDF), which aims to alter what the model believes. Ideally, synthetic documents would be mixed into pre- or mid-training, but every change to a pre-training corpus must be followed by a full post-training run before its effect can be measured, making iteration slow and expensive. Common practice instead applies SDF to an already post-trained model. This is known to leave artifacts and degrade capabilities, and, as we show, it makes the model treat fabricated entities unrelated to the documents as real, a failure we call reality drift. We propose grafting: train the SDF adapter on the pre-trained checkpoint, then add the learned weight update to the post-trained model, which approximates the faithful approach while reusing the existing post-training. We demonstrate this by installing false facts, training misaligned model organisms and applying a constitutional mid-training intervention, across model families up to 284B parameters. Grafting installs the target belief as strongly as SDF on the post-trained model while reducing both reality drift and the loss of preference coherence by more than half on average, and it stays closer to a faithful mid-training run. Because grafting requires no post-training, the same adapter can be applied to any later checkpoint, enabling researchers to iterate quickly on pre-training interventions at the cost of a single fine-tuning run.

Peter Nutter, Dani Roytburg, Clément Dumas +2

ETH Zurich · MATS · Carnegie Mellon University +2

Fine-Tuning LLM Alignment Belief Updating in LLMs Causal Interventions in Language Models

Sep 30, 2026cs.AI

Robust Nash Alignment under Preference Uncertainty

Preference-based alignment methods typically optimize against a single preference model, and can therefore be brittle when pairwise preferences are uncertain: noisy, heterogeneous, or shift after deployment. To address these issues, we propose Robust Nash Alignment, a game-theoretic framework for alignment to uncertain pairwise preferences. Our formulation has a major learner seeking a policy with a large worst-case win rate against both an adversarial competitor and any preference kernel lying in an ambiguity set around a nominal preference. When the ambiguity set captures the uncertainty in preferences, the resulting robust objective of the game directly yields a certified lower bound on worst-case performance. However, we note this problem is computationally challenging to optimize, and to address this, we introduce a four-player primal-dual proxy game involving the leader policy, follower policy, adversarial kernel, and dual variable, and develop a single-loop optimistic mirror descent-ascent algorithm for it. We show that the proxy always lower-bounds the truncated hard-constrained objective, quantify the proxy-to-hard gap, and characterize an exactness condition under which the proxy recovers the robust objective. We then prove an

\mathcal{O}(1/\sqrt{T})

average-iteration convergence for the proxy-game duality gap, which implies a near-optimal robust policy for the original robust objective. Experiments on controlled tabular games and LLM alignment with uncertain preference further validate the convergence theory and show improved performance over nominal baselines.

Shihab Ahmed, Debamita Ghosh, David Tang +3

University of Central Florida · Arizona State University · University of Colorado Boulder

LLM Alignment Preference Alignment Robust RL Nash Equilibrium Computation

Sep 30, 2026cs.CL

Emergent Unfaithfulness: How Alignment Training Causes Language Models to Silently Override Task Faithfulness

Large language models are characterized by three key properties: capability, alignment, and faithfulness. Prior work studies the tradeoffs between capability and alignment, and between capability and faithfulness, but a third tension remains underexplored: the alignment-faithfulness conflict. We show that aligned models systematically deviate from their inputs on unsafe or sensitive content without disclosing the modification, a failure mode we call alignment-induced unfaithfulness (AIU). Unlike capability-driven unfaithfulness, which comes from errors in knowledge or reasoning, this is induced by post-training mechanisms that override adherence to the input. We introduce FaithConflict, a controlled dataset isolating both conflicts, and two complementary taxonomies: behavioral (B1-B8) and chain-of-thought reasoning (C0-C6). Across models, AIU increases with scale and more sharply than capability-driven unfaithfulness, a reverse scaling law; intermediate checkpoints show it is amplified during post-training, with DPO the stage at which the gap both grows most and becomes least visible. Prompting-based mitigation does not resolve it, revealing a capability-alignment-faithfulness trilemma in the design and evaluation of LLMs.

Pardis Sadat Zahraei, Janvijay Singh, Gokhan Tur +1

University of Illinois Urbana-Champaign

LLM Alignment Instruction Following

Sep 30, 2026cs.AI

Values as Style: Disentangling Values from Semantics with One-Way Mixing for Low-Damage LLM Steering

Value steering should change an LLM's normative priorities while preserving the scenario, facts, and task constraints underlying its answer. Conventional activation edits often change both. We introduce an editable semantic-value interface on frozen residual states, with a one-way semantic-to-value pathway that grounds value recognition in context. Stop-gradient blocks feedback through this pathway; swap consistency, topic de-confounding, and decorrelation encourage selective codes. At inference, editing the value code produces a residual delta while holding the semantic code fixed. On two instruction-tuned backbones, this interface improves semantic preservation and reduces benign refusals at comparable value alignment. A matched mixing-by-gating ablation separates representation learning from selective edit activation, and dimension-matched probes establish improved code selectivity. Against validation-selected prompting on LLaMA-3.1-8B, the method achieves comparable alignment (0.750 vs. 0.748), higher BERTScore (0.938 vs. 0.923), and fewer contradictions (5.1% vs. 7.6%). Human ratings and cross-taxonomy controls provide complementary evidence for low-damage value steering.

Jiale Dai, Hongcan Deng, Liuxian Ma +2

State Key Laboratory of General Artificial Intelligence, School of Intelligence Science and Technology, Peking University · University of Chinese Academy of Sciences · College of Artificial Intelligence, Tsinghua University

Disentangled Representation Learning LLM Alignment Language Model Steering Causal Interventions in Language Models

Sep 30, 2026cs.CL

Ready2Blend: From Natural-Language Instructions to Composable Alignment Prompts

Continual alignment requires LLMs to adapt to new requirements without forgetting previously acquired behaviors. Natural-language instructions are flexible and composable but offer only indirect control, whereas post-training provides stronger adaptation at the cost of repeated parameter updates. We introduce Ready2Blend, which combines the flexibility of natural language with learned alignment. AlignFormer maps each requirement to a fixed-length alignment prompt stored in a modular prompt bank, while the backbone and prior prompts remain frozen. Composability regularization transfers the semantic geometry of textual requirements into prompt space, enabling inference-time blending and reweighting. Across two practical continual alignment settings, Ready2Blend is the only frozen-backbone method that matches post-training-based alignment methods, reaching

93.1

-

98.5\%

of a joint-training reference with competitive retention, while requiring only a few prompt tokens and up to

4.3\times

less training time. Its modular design further enables weighted personalization and order-free composition without retraining. Code will be released upon acceptance.

Jeesu Jung, Hwan Chang, Juseon Do +5

Korea Advanced Institute of Science and Technology · Samsung SDS

Continual Learning for LLMs LLM Alignment Prompt Tuning Inference-Time Language Model Alignment Prompt Engineering

Sep 30, 2026cs.AI

Multi-LLM Collaborative Alignment via Stackelberg Games

A pool of language models can collaborate and improve collectively by learning from one another's responses. These interactions depend on the instructions used during training. Existing methods typically sample instructions uniformly, even though their usefulness may change as the models improve: an instruction on which models' responses once differed in quality may later be answered equally well, while a previously difficult instruction may begin to provide a useful learning signal. We propose Stackelberg Alignment, a game-theory-inspired leader-follower framework that turns instruction selection into an adaptive curriculum. An EXP3 bandit acts as the leader, allocating a fixed sampling budget across instructions and updating its sampling distribution using a reward that combines instruction difficulty and response discriminability. The language models act as followers: they respond to the selected instructions, evaluate one another's responses, and learn from the resulting preference signals through DPO or GRPO. The framework uses Elo-style reputation-weighted peer judgment and reputation-based opponent matching to support reliable and competitive model interactions. Experiments across three heterogeneous model pools and 12 benchmarks spanning scientific discovery, reasoning, code, instruction following, and knowledge show that Stackelberg Alignment achieves the highest macro-average across three diverse model pools, outperforming the strongest training-time baseline by up to 7.4% and the best static inference baseline by 12-25%. Analysis confirms that the adaptive leader concentrates duels on the most informative instructions, and ablations show that both reputation-weighted judgment and reputation-based matching improve the effectiveness of multi-LLM evolution.

Christina Hahn, Shangbin Feng, Dean Light +3

University of Washington · Amazon · University of British Columbia

RL for Language Models LLM Alignment Multi-LLM Collaboration LLM Training

Sep 30, 2026cs.LG

Optimal Design for Active Preference Learning with Biased LLM Judges

Learning from human preferences is central to large language model (LLM) alignment, but human preference annotation is costly. Active preference learning reduces this cost by selecting informative comparisons, and LLM judges can provide additional scalable feedback. However, the preferences of the judges may deviate from those of the target human population. Even after calibration on trusted reference data, active acquisition can shift the comparison distribution and expose residual judge bias. We therefore incorporate judge deviations into the acquisition design rather than relying on a separate calibration stage. Under joint estimation, comparisons that appear highly informative about the reward may also reflect judge bias and therefore provide less information about human preferences. To address this issue, we propose Nuisance-Adjusted Optimal Design (NAOD), a comparison-selection strategy that prioritizes policy-relevant target information after nuisance adjustment and uses the Frank-Wolfe algorithm for optimization. Theoretically, we establish a sharp conditional local asymptotic minimax lower bound on policy risk and construct an estimator that attains it. We further characterize the finite-sample cost of learning the nuisance representation and show that representation error can reverse an oracle design advantage. Finally, we validate these predictions experimentally and evaluate NAOD on Chatbot Arena data across 17 judges, 15 budget configurations, and 15 random cluster-level splits. NAOD reduces the mean regret of proxy policy by 29.1% relative to a matched target-information design, outperforms existing methods, and improves human-preference prediction on held-out data.

Zhongman Du, Huiming Zhang, Haodong Zhu +1

Beihang University · Zhongguancun Academy · Hangzhou Innovation Institute of Beihang University

Pairwise Preference Learning LLM Alignment Active Learning

Sep 29, 2026cs.CL

Beyond Oracle Communication: Benchmarking Interactive Intent Alignment Under Miscommunication and Evolving User Intent

Modern LLM agents increasingly tackle complex tasks through interactive, long-horizon exchanges with users, while existing benchmarks generally assume that users always accurately and sufficiently communicate a fixed intent. However, this oracle communication assumption rarely holds in practice: users may miscommunicate, change their goals, and run out of patience. We define this task setting as Interactive Intent Alignment, where agents must recover and continuously track the user's current intent despite imperfect communication and evolving goals. To study this setting, we introduce Drift-Bench++, a principled benchmark construction pipeline for verified executable tasks with controlled misalignment and intent shifts, along with an interaction protocol featuring finite patience, diverse simulated users, and silent interaction-conditioned shifts. We further develop GRIP, a comprehensive evaluation protocol covering task grounding, user realism, inquiry effectiveness, and adaptation to evolving intent. Across diverse environments, models, and interaction conditions, stronger interaction consistently helps but remains far from oracle performance; Validation on deployed ProdAgent sessions further shows that the modeled failures are prevalent and consequential in deployment. By providing a unified, executable benchmark for interactive intent alignment, Drift-Bench++ offers a foundation for evaluating and advancing agents under realistic communication and evolving intent.

Zheyuan Zhang, Mengyuan Chao, Ke Xiao +5

ByteDance Inc., USA · University of Notre Dame

LLM Alignment LLM Agent Evaluation AI Agent Benchmarks Long-Horizon LLM Agents

Sep 29, 2026cs.AI

Aligned Data Can Induce Misalignment via Context Confusion

Large language models (LLMs) are frequently updated for various use cases, where filtering out misaligned training samples is a common practice for preventing post-update misalignment. However, alignment is inherently context-dependent: a recommendation that is aligned in one context may be inappropriate in another. For example, in response to the question "What should a researcher do with the research data?", recommending that the researcher preserve the data for reproducibility is aligned. In contrast, recommending data saving in response to "What should a mobile-app developer do with users' sensitive data?" may be inappropriate from a privacy perspective. Starting from this observation, we identify a post-training phenomenon where aligned training induces misaligned behavior in other contexts. We call this phenomenon context confusion. We demonstrate context confusion across three domains: (1) Gender Equality, (2) Privacy, and (3) Physical Safety. We further show that context confusion causes narrow misalignment, in contrast to emergent misalignment, and is not effectively reduced by injecting general alignment data, but can be substantially reduced by including targeted alignment data for the misaligned domain or providing in-context learning examples during inference. Lastly, we provide a mechanistic explanation of context confusion. We observe that queries from different domains can undergo similar representational shifts during the fine-tuning. Consequently, a query from a different domain may activate the same behavioral feature learned during fine-tuning, which causes the behavior to transfer to a context where it is misaligned. Based on our findings, we argue that it is difficult to predict the alignment state of a model after training by inspecting the training data alone, which highlights the importance of comprehensive post-training alignment evaluations.

Yavuz Bakman, Duygu Nur Yaldiz, Baris Askin +4

University of Southern California · Carnegie Mellon University · Amazon +1

LLM Alignment LLM Post-Training Inference-Time Language Model Alignment

Sep 29, 2026cs.CL

The Unequal Influence of Bad Advice: Using Training Data Attribution to Modulate Emergent Misalignment

Fine-tuning large language models on narrow, misaligned tasks can undo their post-training alignment and induce novel misaligned behaviors -- a phenomenon known as \emph{emergent misalignment} (EM). EM has been linked to persona-like representations, where fine-tuning might reduce loss by amplifying a harmful or 'evil' persona. It remains unclear which properties of the training data drive this effect: whether all harmful examples contribute approximately equally to misalignment and whether different models are equally affected by the same fine-tuning examples. In this work, we use training data attribution to quantitatively estimate how much each harmful example contributes to EM. We benchmark the quality of the attribution via retraining -- a sound attribution score should enable us to enhance or attenuate EM by filtering data on that score. Score-based filtering can substantially enhance or attenuate EM; we find that both data-attribution scores and a black-box harmfulness score can identify consequential examples. All models we test become misaligned when trained on the same dataset, and influence scores perform best when filtering data from the same model that computed them. We find cross-model generalization of influence scores from scores derived from the three model families we tested, but this generalization does not recover same model filtering performance.

Gonçalo Paulo, Louis Jaburi, Nora Belrose +2

EleutherAI

LLM Alignment LLM Fine-Tuning Emergent Misalignment in Language Models Training Data Selection Training Data Attribution

Sep 29, 2026cs.CL

Correct, Don't Delete: Mitigating Emergent Misalignment with Corrective Supervision

Fine-tuning a language model on a narrow set of harmful demonstrations, such as bad medical advice, can make it broadly misaligned on unrelated questions, a phenomenon known as emergent misalignment (EM). The usual defense is to find the offending rows and delete them, but a row locator failed our held-out test and deleting rows helps less than expected. We ask a different question: given a fixed set of poisoned rows, is it better to correct them than to remove them? We fine-tune Qwen2.5-14B-Instruct on a mixture of bad medical advice and benign chat data, select a quarter of the poison rows in advance, and either delete them or replace each with a corrected answer to the same prompt, keeping everything else the same. Replacing the rows cuts the EM rate by about a third and improves answers on held-out medical questions, while deleting the same rows has little measurable effect. The advantage is larger when half the poison rows are corrected, and it holds on a second base model and a second misaligned model organism. The content of the replacement appears to matter: paraphrasing the rows while keeping their bad advice shows no clear benefit, and the correct answers distributed with the dataset appear to do about as well as our rewriter's. Realigning an already-poisoned model with further fine-tuning is known to work, but which data does the work has not been compared directly. We find that a short round of training on corrections beats the same amount of training on generic chat data, that corrections on other medical prompts do roughly as well as corrections of the poisoned prompts themselves, and that instructing the correction writer to model a careful, harm-avoiding assistant adds no measurable benefit over plain corrections. In the settings we tested, correcting harmful training data reduces EM more than deleting it.

Jacob Epifano

Independent researcher

LLM Alignment Emergent Misalignment in Language Models Data Contamination in Language Models

Sep 29, 2026cs.CL

ContextAdapt: Evaluating Contextual Adaptation and Value Alignment in LLMs

Values such as honesty, autonomy, and confidentiality are often regarded as general principles underpinning AI alignment. However, what it means to act in accordance with these values can depend on the context in which a decision is made. In this paper, we ask whether large language models (LLMs) appropriately adapt the application of a value across professional settings, while remaining consistent when contextual changes do not alter the relevant professional norm. To study this, we introduce ContextAdapt, an evaluation framework covering honesty, autonomy, and confidentiality across medicine, law, finance, and national security. Drawing on primary-source professional and regulatory documents, we construct a value x domain framework and use this to develop scenarios testing both default professional rules and recognised exceptions. We evaluate 12 LLMs on both the actions they recommend and the justifications they provide. In our main experiment, models achieve 95.6% mean appropriateness, although the use of the correct domain-specific justification varies substantially across models, from 25.6% to 76.9%. In a separate factorial experiment, explicitly naming the professional domain and changing the role of the model have limited effect on behaviour. Varying stakes, however, reveals severe but localised failures: in some cases, models alter their responses even though the underlying professional obligation remains unchanged. In particular, perceived severity appears to act as a cue for disclosure across both honesty and confidentiality scenarios. These results show that evaluating value alignment requires us to consider not only whether models follow abstract principles, but whether they apply them appropriately across different contexts.

Olivia Macmillan-Scott, Mirco Musolesi

Centre for AI, Department of Computer Science, University College London · Department of Computer Science and Engineering, University of Bologna

LLM Alignment Language Model Safety Evaluation

Sep 28, 2026cs.LG

FLOORA: A Human-Aligned Domain-Specific Language Model for Architectural Design

Foundation models are powerful generators, but many engineering domains require structured representations that general-purpose systems handle poorly. We introduce FLOORA (Floor Layout Optimization with RL Alignment), a family of small domain-specific language (DSL) models for architectural layout generation. With specialized data and alignment, our 0.6B model outperforms much larger frontier models, achieving VLM judge win rates up to 92.0% on out-of-distribution real-world buildings and 96.0% on synthetic buildings. Human evaluations further corroborate these results, with FLOORA selected as the best model in 89.3% of evaluations. FLOORA combines a token-efficient DSL, custom tokenization, domain-specific pretraining, supervised fine-tuning (SFT), and reinforcement learning (RL) with learned human-preference and verifiable rewards. This pipeline improves architectural and geometric validity, supported by extensive empirical evaluation and ablation studies. Although focused on architecture, our results suggest that similar domain-specific recipes may be useful in other engineering domains with structured, verifiable outputs. Datasets, models, and inference code are available at https://github.com/AutodeskAILab/floora.

Sahand Rezaei-Shoshtari, Patryk Wozniczka, Shu Ishida +14

Autodesk Research

LLM Alignment Language Modeling

Sep 28, 2026cs.CL

From Normative Frameworks to Alignment Data: Constructing and Evaluating SFT and Preference Data

Aligning language models with a specified normative framework requires translating abstract principles into concrete examples and preference signals from which models can learn. We present an expert-driven methodology for constructing such alignment data and apply it to a normative framework grounded in Islamic ethical, theological, and jurisprudential traditions. Over approximately one year, seven domain experts systematically probed language models to identify alignment deficiencies, curated desired responses, and constructed preference pairs from model outputs and expert judgments. The resulting Arabic-English datasets contain approximately 2.8K supervised fine-tuning (SFT) examples and 5.4K preference pairs spanning a broad range of normative domains. We evaluate the datasets through controlled post-training experiments comparing a Baseline model with models incorporating the curated SFT data alone and both the SFT and preference data. In blind expert evaluation on 150 separately constructed prompts, the model trained with the curated SFT data was preferred over the Baseline in 51.3% of assessor judgments, compared with 14.4% in the opposite direction (p < .001 at the prompt level). Adding the preference data resulted in a smaller difference, with the model trained with both datasets preferred over the SFT model in 28.0% of judgments versus 20.9% in the opposite direction; this difference was not statistically significant at the prompt level (p = .166). Standard Arabic and English benchmarks show no broad degradation in general-purpose capabilities. These results demonstrate how expert-defined normative principles can be systematically operationalized into alignment data and evaluated through controlled model training.

Husrev Taha Sencar, Rezart Beka, Danish Naeem +6

Qatar Computing Research Institute, HBKU, Qatar. · College of Islamic Studies, HBKU, Qatar. · Argumentation and Conflict Studies, Ibn Haldun University, Turkiye. +1

LLM Alignment Preference Alignment

Sep 28, 2026cs.LG

BA-DPO: Bias-Adjusted Direct Preference Optimization for Language Model Alignment

Preference-based alignment methods such as Direct Preference Optimization (DPO) use pairwise preferences labeled by human annotators to fine-tune language models. However, annotators carry systematic biases toward some attributes: a name that signals a gender or an ethnicity, a persona, a language variety, a formatting convention, or length. If not properly addressed, these systematic biases can be absorbed and amplified during alignment. Existing methods address length bias or annotator disagreement, but fail to eliminate biases toward arbitrary attributes. To address this limitation, we propose Bias-Adjusted DPO (BA-DPO), a generalization of DPO that adds one bias parameter per annotator toward responses carrying a declared attribute. We prove that the objective is convex in the bias parameters and that the votes identify each annotator's bias up to a shared constant. The remaining constant is what fixes the aligned model's attribute rate: by default the rate of the reference model, or a target rate, which we use to bring a biased policy to statistical parity. On a corpus with planted biases, DPO drives the attribute from a balanced start to probability 0.96 and BA-DPO removes 81 to 95% of that shift; on MultiPref with real annotators it removes about half of DPO's lengthening. Both hold at 0.5B with full fine-tuning and at 8B with LoRA, at no higher KL than DPO and no loss in judged quality.

Antonio Ferrara, Alberto Rumi, Francesco Bonchi

Intesa Sanpaolo AI Research

LLM Alignment Debiased ML Direct Preference Optimization Demographic Bias in Language Models

Sep 28, 2026cs.LG

ABC-Align: Prediction-Powered Alignment with Adaptive Bias Control

Language model post-training is often bottlenecked by the need for human-collected preference data, which is expensive and difficult to scale. Reinforcement learning from AI feedback (RLAIF) style approaches that leverage pseudo labels offer an abundant alternative but introduce systematic biases that degrade downstream alignment. Recent general-purpose semi-supervised methods correct for teacher bias using a small set of human-labeled examples, but suffer from high variance especially when human annotations are scarce. To this end, we propose ABC-Align, leveraging abundant pseudo label signal to minimize variance and applying a lightweight, adaptive correction grounded in the human-labeled subset. The correction strength is tuned automatically during training using plug-in estimates of the relevant bias--variance quantities. On LLM alignment with RLHF, DPO, and GRPO where human feedback is scarce, we empirically demonstrate that ABC-Align achieves superior performance over prior semi-supervised baselines in a series of experiments on an increasing scale. Our code is available at https://github.com/SewoongLab/abc-align .

Eric Frankel, Banghua Zhu, Sewoong Oh +1

University of Washington, Seattle

RL for Language Models LLM Alignment Language Model Post-Training Prediction-Powered Inference

Sep 27, 2026cs.CL

When Do Agents Help? Embedding, LLM and Agentic Alignment of Classical Texts and Their Translations

Classical texts aligned with their translations support machine translation, retrieval and computational research, but evidence comparing alignment workflows is scattered. This study compares seven systems on 452 texts in Pali, Sanskrit, Mishnaic Hebrew and Tibetan, comprising 9,833 human-aligned units: four embedding pipelines, a direct LLM call, an autonomous agent, and the agent revised by an independent auditor. Generative workflows recover 93-94% of reference correspondences, against at most 77% for embeddings. A ceiling analysis shows that sentence boundaries make some references unrepresentable by the embedding pipelines. Reference recovery is similar across generative workflows: the agent's advantage is 0.5 percentage points (95% CI -0.02 to 1.17), and auditing adds no established benefit. Agents nevertheless produce structurally valid output for all 452 texts, against 437 for direct calls. A blinded three-LLM panel assesses every generative mismatch against the source and human reference. Most mismatches are labelled defensible editorial variation; consensus major-error labels cover only 0.06-0.14% of units. The panel labels significantly fewer residual defects for agents than direct calls (0.7% versus 1.4%), suggesting that reference recovery alone understates alignment quality. On ten long Pali discourses taken as published online, agents and audited agents raise recovery from the direct call's 71% to 84% and 92%. Identical reference-located chunks bring all three to 93%. Agents thus improve structural reliability and reduce judged defects on short passages, while their large recovery advantage on long documents disappears after chunking. In this setting, independent auditing offers little measurable additional benefit on prepared passages.

Máté Metzger

Independent Researcher

LLM Alignment LLM Agent Evaluation LLM Agent Reliability Sequence Alignment

Sep 27, 2026cs.LG

Direct Hidden-State Alignment: Mapping and Controlling Preference Expression in LLMs

In many settings, post-training need not create the target behavior from scratch: the base model can already produce it, but not reliably. This shifts part of preference alignment from capability acquisition to behavioral expression. We ask how a specified preference is represented in native model computation, what prevents target-supporting computation from reliably dominating generation, and whether this structure can directly guide control. We introduce Residual Competition Maps (RCMs), which map a behavioral preference onto signed causal effects of native residual computation. Across preference domains, RCMs reveal coexisting target-supporting and target-competing effects, input-dependent component roles, and cases where a single native-component intervention reverses the preference outcome. DPO substantially reorganizes these effects and can weaken opposition without guaranteeing its removal. We then propose Direct Hidden-State Alignment (DHSA), which treats inference-time hidden states rather than base-model weights as the direct adaptation space. RCM-guided Causal Activation State Transition (CAST) implements DHSA through local state interventions at a small number of preference-relevant interfaces while freezing the base model. With only 256-16,384 controller parameters, CAST reaches DPO-competitive operating points across three preference domains, can complement DPO-trained models, and can be enabled or removed at inference time.

Fansheng Zhang, Shengran Guo, Zexiao Wang +3

Chengdu University · North Carolina State University · Fudan University +2

LLM Alignment Preference Alignment Inference-Time Language Model Alignment Causal Interventions in Language Models Activation Steering

Sep 24, 2026cs.AI

Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures

Detectors of alignment failures screen deployed language models and score alignment benchmarks. Most are generative judges that spend a decoding pass on every criterion, and classifiers that read token probabilities, such as Llama Guard, still score one fixed label per call. Jev, a model trained with reinforcement learning for calibrated decisions (RLCD), answers many typed questions about one input with calibrated probabilities in a single call. Whether it detects alignment failures has not been measured. We present RLCDAlignBench, which benchmarks Jev on ten alignment failures: sycophancy, jailbreaks, deception, prompt injection, hallucination, privacy violation, social bias, reward hacking, concealing uncertainty, and power seeking. It spans 44 benchmarks and five target models, labelled by each benchmark's scorer and, on two, by humans. Many of these failures are relational, defined against a reference, such as the user's belief or an injected instruction, that the response alone does not reveal. Our key idea is therefore to vary what Jev is asked separately from what it sees: the question's wording and answer type on one side, the fields of the input on the other. A single generic question reaches a median AUROC of 0.886 zero-shot and beats supervised baselines on most benchmarks. Question wording matters little, while context matters more, mostly through fields that encode the label. Jev matches the reference scorer's agreement with human labels, surfaces label defects in existing benchmarks, and costs 63x less than LLM-judge scorers. Code and data: https://github.com/sumleo/RLCDAlignBench.

Ruoqi Guo, Yi Liu, Gelei Deng +6

Griffith University · Nanyang Technological University · UNSW +4

LLM Safety Benchmarks LLM Alignment Reinforcement Learning