Via Negativa for AI Alignment: Why Negative Constraints Are Structurally Superior to Positive Preferences

• URL: http://arxiv.org/abs/2603.16417v1
• Date: Tue, 17 Mar 2026 11:52:18 GMT
• Title: Via Negativa for AI Alignment: Why Negative Constraints Are Structurally Superior to Positive Preferences
• Abstract summary: Training large language models with negative-only feedback can match or exceed standard reinforcement learning from human feedback.<n>No unified theoretical account explains why negative signals are so effective.<n>Positive preferences and negative constraints are structurally asymmetric.
• Score: 0.786119752211706
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent empirical results have demonstrated that training large language models (LLMs) with negative-only feedback can match or exceed standard reinforcement learning from human feedback (RLHF). Negative Sample Reinforcement achieves parity with PPO on mathematical reasoning; Distributional Dispreference Optimization trains effectively using only dispreferred samples; and Constitutional AI outperforms pure RLHF on harmlessness benchmarks. Yet no unified theoretical account explains why negative signals are so effective. This paper proposes such an account: positive preferences and negative constraints are structurally asymmetric. Positive preferences ("which is better") encode continuously coupled, context-dependent human values that cannot be exhaustively specified -- leading models to learn surface correlates such as agreement with the user (sycophancy). Negative constraints ("what is wrong") encode discrete, finite, independently verifiable prohibitions that can converge to a stable boundary. This asymmetry -- rooted in Popper's falsification logic and the epistemology of negative knowledge -- explains both the sycophancy failure of preference-based RLHF and the surprising effectiveness of negative-signal methods. We argue that alignment research should shift its center of gravity from "learning what humans prefer" to "learning what humans reject," and offer testable predictions for this framework.

Related papers

• When Hard Negatives Hurt: Bridging the Generative-Discriminative Gap in Hard Negative Synthesis for Retrieval [45.5843471557695]
  We show that naively incorporating generated negatives into contrastive learning often degrades retrieval performance.<n>Our analysis reveals two compounding failure modes: discriminative-agnostic generation and source-dependent shortcuts.<n>To close this gap, we propose CausalNeg consisting of two main modules.
  arXiv  Detail & Related papers  (2026-05-31T15:48:30Z)
• ResRL: Boosting LLM Reasoning via Negative Sample Projection Residual Reinforcement Learning [60.411097231348975]
  Reinforcement Learning with Verifiable Rewards (RLVR) enhances reasoning of Large Language Models (LLMs)<n>This paper proposes negative sample projection Residual Reinforcement Learning (ResRL) that decouples similar semantic distributions among positive and negative responses.
  arXiv  Detail & Related papers  (2026-05-01T03:57:44Z)
• ReNCE: Learning to Reason by Noise Contrastive Estimation [7.590073864595161]
  GRPO is a standard approach to endowing pretrained LLMs with reasoning capabilities.<n>We propose an explicit contrastive learning approach. Instead of estimating advantages, we bifurcate $K$ outcomes into positive and negative sets.
  arXiv  Detail & Related papers  (2026-01-30T00:57:31Z)
• Correct and Weight: A Simple Yet Effective Loss for Implicit Feedback Recommendation [36.820719132176315]
  This paper introduces a novel and principled loss function, named Corrected and Weighted (CW) loss.<n>CW loss systematically corrects for the impact of false negatives within the training objective.<n> experiments conducted on four large-scale, sparse benchmark datasets demonstrate the superiority of our proposed loss.
  arXiv  Detail & Related papers  (2026-01-07T15:20:27Z)
• Unregularized Linear Convergence in Zero-Sum Game from Preference Feedback [50.89125374999765]
  We provide the first convergence guarantee for Optimistic Multiplicative Weights Update ($mathtOMWU$) in NLHF.<n>Our analysis identifies a novel marginal convergence behavior, where the probability of rarely played actions grows exponentially from exponentially small values.
  arXiv  Detail & Related papers  (2025-12-31T12:08:29Z)
• Negative-Prompt-driven Alignment for Generative Language Model [34.191590966148816]
  We propose NEgative-prompt-driven AlignmenT to guide language models away from undesirable behaviors. NEAT explicitly penalizes the model for producing harmful outputs, guiding it not only toward desirable behaviors but also steering it away from generating undesirable, biased responses. Extensive experiments validate NEAT's effectiveness in significantly enhancing language models' alignment with human values and preferences.
  arXiv  Detail & Related papers  (2024-10-16T03:30:09Z)
• Negating Negatives: Alignment with Human Negative Samples via Distributional Dispreference Optimization [37.8788435790632]
  Large language models (LLMs) have revolutionized the role of AI, yet pose potential social risks. Existing methods rely on high-quality positive-negative training pairs, suffering from noisy positive responses that are barely distinguishable from negative ones. We propose Distributional Dispreference Optimization (D$2$O), which maximizes the discrepancy between dispreferred responses and the generated non-negative ones.
  arXiv  Detail & Related papers  (2024-03-06T03:02:38Z)
• Making Large Language Models Better Reasoners with Alignment [57.82176656663245]
  Reasoning is a cognitive process of using evidence to reach a sound conclusion. Recent studies reveal that fine-tuning LLMs on data with the chain of thought (COT) reasoning process can significantly enhance their reasoning capabilities. We introduce an textitAlignment Fine-Tuning (AFT) paradigm, which involves three steps.
  arXiv  Detail & Related papers  (2023-09-05T11:32:48Z)
• Language Model Pre-training on True Negatives [109.73819321246062]
  Discriminative pre-trained language models (PLMs) learn to predict original texts from intentionally corrupted ones. Existing PLMs simply treat all corrupted texts as equal negative without any examination. We design enhanced pre-training methods to counteract false negative predictions and encourage pre-training language models on true negatives.
  arXiv  Detail & Related papers  (2022-12-01T12:24:19Z)
• Learning to Estimate Without Bias [57.82628598276623]
  Gauss theorem states that the weighted least squares estimator is a linear minimum variance unbiased estimation (MVUE) in linear models. In this paper, we take a first step towards extending this result to non linear settings via deep learning with bias constraints. A second motivation to BCE is in applications where multiple estimates of the same unknown are averaged for improved performance.
  arXiv  Detail & Related papers  (2021-10-24T10:23:51Z)
• Deconfounding Scores: Feature Representations for Causal Effect Estimation with Weak Overlap [140.98628848491146]
  We introduce deconfounding scores, which induce better overlap without biasing the target of estimation. We show that deconfounding scores satisfy a zero-covariance condition that is identifiable in observed data. In particular, we show that this technique could be an attractive alternative to standard regularizations.
  arXiv  Detail & Related papers  (2021-04-12T18:50:11Z)
• NPCFace: Negative-Positive Collaborative Training for Large-scale Face Recognition [78.21084529159577]
  We study how to make better use of hard samples for improving the training. The correlation between hard positive and hard negative is overlooked, and so is the relation between the margins in positive and negative logits. We propose a novel Negative-Positive Collaboration loss, named NPCFace, which emphasizes the training on both negative and positive hard cases.
  arXiv  Detail & Related papers  (2020-07-20T14:52:29Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.