Pith. sign in

REVIEW 3 major objections 4 minor 34 references

Improving Consistency in Vehicle Trajectory Prediction Through Preference Optimization

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Fine-tuning trajectory predictors on automatically ranked joint futures, using an adapted SimPO preference-optimization loss, substantially reduces scene collision rates while barely hurting accuracy.

desk verdict First preference-optimization paper for trajectory prediction, with a sensible SimPO adaptation and promising collision-rate numbers, but the reported gains are entangled with the order-pairing heuristic used in both training and evaluation. read the letter →

arxiv 2507.02406 v1 pith:ZBZDI5JN submitted 2025-07-03 cs.LG

Caio Azevedo , Lina Achaji , Stefano Sabatini , Nicola Poerio , Grzegorz Bartyzel , Sascha Hornauer , Fabien Moutarde This is my paper
classification cs.LG
keywords trajectorypredictionpreferenceoptimizationsceneconsistencycollisionrateSimPOPlackett-Lucerankingmulti-agentfine-tuning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether preference optimization, the technique used to align large language models with human preferences, can make multi-agent trajectory predictors internally consistent. It answers yes: by automatically ranking a model's predicted joint futures from best to worst according to a collision-penalizing cost, then fine-tuning with an adapted SimPO loss based on Plackett-Luce ranking, the likelihood of collision-containing modes drops sharply. On Argoverse 2, QCNet's scene collision rate falls 27% and its probability-weighted collision rate falls 57%, while its accuracy metric degrades only 1%. The method adds no inference-time cost and works on both marginal and joint predictors.

What carries the argument

The machinery is an adapted SimPO objective for multi-agent trajectory prediction. Agent-level predictions are paired by likelihood order into $K$ scene modes, and the scene-level log-likelihood is the average of the agents' logits in that mode. The reward becomes $r_\theta(x,y) = \beta \log \pi_\theta(y|x)$ with no length normalization, and the loss is the Plackett-Luce ranking loss with a rank-scaled margin $k\gamma$, so higher-ranked modes must beat lower-ranked modes by an increasing margin. The preference ranking itself comes from an automatic cost combining average per-agent final displacement error with the MotionDiffuser repeller cost, which penalizes pairwise agent distances below a threshold. This loss is applied only on a subset of training scenes that contain collisions or large cost spreads.

What would settle it

Take a validation scene where the model's top individual modes cross-collide in every order-based pairing, and compare the probability-weighted collision rate under the paper's pairing with the rate under a randomized or learned pairing of the same agent trajectories; if the improvement disappears or inverts under an alternative pairing, the reported gains are an artifact of the aggregation heuristic rather than of preference optimization.

Watch

Extended reading notes

Core claim

The central claim is that scene consistency in trajectory prediction can be improved by treating joint modes as ranked outputs and optimizing their likelihoods directly, without retraining the architecture or adding inference overhead. The authors adapt SimPO: the reward of a mode is its aggregated log-likelihood $\beta \log \pi_\theta(y|x)$, the loss uses Plackett-Luce over all $K$ modes, and the target reward margin scales with rank position ($k\gamma$). Modes are ranked automatically by a cost combining average final displacement error with a repeller collision cost, and collision-free modes are pushed up in probability while collision modes are pushed down. Across Argoverse 2, INTERACTION, and the Waymo Open Motion Dataset, the fine-tuning reduces collision rates for QCNet, FJMP-Marginal, FJMP-Joint, and BeTop while MinJointFDE degrades by only a few percent; oversampling to $K=15$ pushes QCNet's scene collision rate reduction to 64%.

Load-bearing premise

The load-bearing premise is that pairing each agent's k-th most likely trajectory into the k-th scene mode, then averaging the agents' logits, gives a faithful enough joint likelihood; if that pairing is unrepresentative of the true joint distribution, the preference loss optimizes a synthetic objective that may not reflect the collision rates measured at inference.

Editorial extensions

If this is right

  • Marginal predictors can acquire joint consistency without joint decoding, removing the main reason to switch to heavier joint models.
  • Planners consuming the fine-tuned predictions see fewer collision modes and much lower probability mass on remaining collision modes, making downstream safety filtering easier.
  • Increasing the number of sampled modes before fine-tuning amplifies the effect: QCNet with $K=15$ reaches a 64% scene collision rate reduction, showing diversity is a lever for consistency.
  • Because the ranking metric is pluggable, the same fine-tuning recipe transfers to other definitions of good joint futures with no architectural change.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The order-based pairing heuristic is the main internal choice; a learned joint scoring head or interaction-aware pairing could make the optimized objective match inference more closely and reduce the small accuracy trade-off.
  • The method's dependence on mode diversity implies it will work best on anchor- or intention-based decoders; improving mode coverage in weaker decoders may unlock larger consistency gains.
  • Beyond collisions, the same preference-optimization loop could target rule violations, comfort, or social norms, since any computable ranking can supply the preference signal.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes a fine-tuning method that adapts SimPO preference optimization to multi-agent trajectory prediction. Given a pretrained marginal or joint predictor, the method ranks predicted scene-level modes by a preference cost combining average FDE and a collision repeller term, extracts a subset of interactive training scenes, and fine-tunes the model with a Plackett-Luce variant of SimPO that uses rank-scaled margins. Experiments on Argoverse 2, Interaction, and Waymo Open Motion Dataset show reduced scene collision rate (SCR) and probability-weighted SCR (pSCR) for QCNet, FJMP variants, and BeTop, with only small degradation in MinJointFDE. The paper also studies the effect of oversampling modes and the sensitivity to the target reward margin γ and collision weight λ.

Significance. If the reported improvements are robust, this is a novel and practically relevant application of preference optimization to trajectory prediction: it is training-time only, model-agnostic, and does not add inference cost. The paper is careful to include ablations over γ and λ, to test both marginal and joint models, and to report accuracy trade-offs. The main limitation is that the scene-level likelihood aggregation and the evaluation metrics are both defined through the same order-based pairing of per-agent trajectories, which confounds the claimed consistency gains with the choice of aggregation heuristic.

major comments (3)
  1. [Section IV-C and V-B] The scene-level likelihoods used for both training and evaluation are constructed by pairing each agent's k-th most likely trajectory and averaging logits (Section IV-C). The SimPO loss in Eq. (12) and the metrics SCR/pSCR in Eqs. (13)-(14) are both computed on modes formed by this exact order-based pairing. Since the paper itself states that 'rearranging agent pairings to achieve consistent predictions is the main effect' (Section V-C), the reported SCR/pSCR improvements may be largely an artifact of the fixed pairing convention rather than evidence of model-independent joint consistency. To support the headline claim, the authors should evaluate with an independent aggregation procedure (e.g., random pairings, a different ranking-based pairing, or a jointly decoded set of modes) and show that the fine-tuned model still reduces collisions under that independent evaluation. Without such an experiment, the central claim is not established outside the specific pairing used for both training and evaluation.
  2. [Equation (12)] The rank-scaled margin term kγ is added inside the exponential in the Plackett-Luce loss. This changes the effective reward from β log πθ(yτ(k)|x) to β log πθ(yτ(k)|x) + kγ, but the derivation from Eq. (11) is not given. It is unclear whether this modification preserves the Plackett-Luce likelihood interpretation or whether it is simply a heuristic that happens to work. The authors should provide a formal justification or at least an explicit statement that this is a heuristic adaptation, and show that the results are not sensitive to the exact placement of γ (inside vs. outside the log-softmax).
  3. [Section V-C, Figure 3] The oversampling experiments report SCR improvements for K=15 but the evaluation protocol is not fully specified. When the model outputs 15 joint modes and the metrics are computed for the 'most probable six joint modes', it is unclear whether the top-6 are selected after re-ranking by the scene-level logits under the order-based pairing, or by some other criterion. Since SCR and pSCR depend on which six modes are retained, this choice could affect the reported 64% improvement. The authors should state exactly how the top-6 are chosen in the K=15 experiments.
minor comments (4)
  1. [Section V-A] In Table I, the first row is labeled 'IntFJMP-Marginal'; this is likely a typo for 'FJMP-Marginal' on the Interaction dataset. Please correct the label for consistency.
  2. [Equation (14)] The definition of pSCR uses πθ(yk|x), but it is not explicitly stated that this is the scene-level likelihood obtained by the order-based aggregation of Section IV-C. Please make this dependence explicit at the point of definition.
  3. [Section IV-A] The preference cost Ck in Eq. (9) uses avgFDEk, which requires ground-truth future trajectories. The paper notes that any metric can in principle be used, but it would be helpful to report results with a purely collision-based preference cost (e.g., λ→∞) to demonstrate that the method does not rely on ground-truth knowledge at fine-tuning time.
  4. [Section V-B] The statement that 'the large decrease in pSCR ... suggests that rearranging agent pairings is the main effect' is important and somewhat undermines the interpretation of SCR improvements as trajectory-shape changes. This should be discussed more prominently, and the paper should distinguish between collision avoidance due to re-pairing versus collision avoidance due to actual trajectory deformation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: the preference objective is driven by external ground-truth and collision geometry, and all reported metrics are measured on held-out benchmarks.

full rationale

This paper is an empirical fine-tuning study, not a derivation, and I find no circular step under the stated criteria. The SimPO objective (Eq. 12) is built from preference rankings determined by an external cost (Eq. 9) combining ground-truth FDE and a geometric repeller cost (Eqs. 7-8), not from the evaluation metrics. Results are measured on held-out splits of Argoverse 2, Interaction, and WOMD, and the reported MinJointFDE degradation shows a real trade-off rather than a forced identity. The pSCR metric (Eq. 14) does share the same scene-level likelihoods (order-based pairing and logit averaging) that the loss reshapes, so pSCR gains are expected from the training objective; this is an evaluation-alignment caveat, not a fitted parameter or input being renamed as a prediction. The paper also transparently acknowledges that rearranging agent pairings is a main effect (Sec. V-C), which is a mechanism explanation rather than a concealment. No load-bearing self-citation or imported uniqueness theorem appears. The central claim, that preference optimization can reduce collision rates on held-out scenes, is supported by self-contained empirical evidence against external benchmarks.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The method depends on several hyperparameters fitted per dataset and model, and on a heuristic aggregation of marginal likelihoods into joint modes. No new physical or conceptual entities are introduced. The axioms are standard preference optimization math plus domain assumptions about collision metrics and the validity of likelihood aggregation.

free parameters (7)
  • beta (SimPO reward scale) = 2.0 for all experiments (Section V-A.d)
    Scales the reward in the SimPO loss; chosen per experiment, not derived.
  • gamma (target reward margin) = 5 for QCNet and BeTop, 15 for FJMP-Marginal (Section V-A.d)
    Controls the required reward difference between ranked modes; strongly affects results (Table III).
  • lambda (collision cost weight) = 10^3 for main results; 0 in ablation (Sections IV-A and V-A.d)
    Weights the repeller cost in the preference metric; lambda=0 also works, showing sensitivity.
  • delta (preference dataset extraction threshold) = 2.5 for Argoverse 2, 1.0 for Interaction (Section V-A.d)
    Threshold for selecting scenes with high variance in preference cost; per-dataset tuning.
  • r (repeller distance threshold) = Not specified numerically in the paper
    Hyperparameter in the repeller cost from MotionDiffuser (Equation 7); controls the minimum distance for an interaction but no value is given.
  • Number of SimPO fine-tuning epochs = 5 (Section V-A.d)
    Chosen for all experiments.
  • Learning rate = 10^-5 (Section V-A.d)
    Fixed for all fine-tuning runs.
assumptions (5)
  • standard math Plackett-Luce model as the preference distribution over rankings
    Adopted from preference optimization literature (Section III-B.b) to extend Bradley-Terry to K-mode rankings.
  • standard math Bradley-Terry model for pairwise comparisons
    Base preference model used in SimPO and DPO (Section III-B.b).
  • ad hoc to paper Agent-level likelihoods can be aggregated into scene-level likelihoods by pairing each agent's k-th most likely trajectory and averaging logits
    Introduced in Section IV-C as a straightforward pairing strategy; not justified as an accurate joint model.
  • domain assumption MotionDiffuser's repeller cost correctly detects collisions and penalizes near-collisions
    Used as the collision metric in Equations 7-8; assumed to be a valid proxy for scene inconsistency.
  • domain assumption Ground-truth future trajectories may be used to define preferences during fine-tuning
    The avgFDE term in the preference cost (Equation 9) requires ground truth, which is available at training time but not at inference.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Consistency in Vehicle Trajectory Prediction Through Preference Optimization." pith.science (2026). https://pith.science/paper/ZBZDI5JN

@misc{pith2026250702406,
  author       = {Pith},
  title        = {Pith review of: Improving Consistency in Vehicle Trajectory Prediction Through Preference Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZBZDI5JN}},
  note         = {Machine review of arXiv:2507.02406}
}
read the original abstract

Trajectory prediction is an essential step in the pipeline of an autonomous vehicle. Inaccurate or inconsistent predictions regarding the movement of agents in its surroundings lead to poorly planned maneuvers and potentially dangerous situations for the end-user. Current state-of-the-art deep-learning-based trajectory prediction models can achieve excellent accuracy on public datasets. However, when used in more complex, interactive scenarios, they often fail to capture important interdependencies between agents, leading to inconsistent predictions among agents in the traffic scene. Inspired by the efficacy of incorporating human preference into large language models, this work fine-tunes trajectory prediction models in multi-agent settings using preference optimization. By taking as input automatically calculated preference rankings among predicted futures in the fine-tuning process, our experiments--using state-of-the-art models on three separate datasets--show that we are able to significantly improve scene consistency while minimally sacrificing trajectory prediction accuracy and without adding any excess computational requirements at inference time.

Figures

Figures reproduced from arXiv: 2507.02406 by the authors.

Figure 1
Figure 1. Illustration of our method on a hypothetical scenario. The input scene (I) is processed by a model that outputs unrealistic interactions, such as collisions [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Qualitative example of SimPO effects, K = 6. Ground-truth future is shown in red and past trajectories in dashed light blue. Predictions with the same number belong to the same mode. Modes 3 and 4 are darker to highlight effect of SimPO. Before, there is a collision between agents 1 and 2 at modes 3 and 4 within the red circle. After, no collisions take place as aggregation changed. Unrealistic modes have decreased … view at source ↗
Figure 5
Figure 5. Example of FJMP-Marginal predictions to demonstrate mode collapse [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Qualitative example of the effectiveness of SimPO combined with [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: This is a known problem of decoding multiple futures [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

34 extracted references · 20 canonical work pages

  1. [1]

    Car Accident Statistics for 2025,

    C. Bieber, “Car Accident Statistics for 2025,” 2024, Forbes, [Online]. Available: https://www.forbes.com/advisor/legal/car-accident-statistics/, 2024

    work page 2025

  2. [2]

    Scene transformer: A unified architecture for predicting future trajectories of multiple agents,

    J. Ngiam, V . Vasudevan, B. Caine, Z. Zhang, H.-T. L. Chiang, J. Ling, R. Roelofs, A. Bewley, C. Liu, A. Venugopal et al., “Scene transformer: A unified architecture for predicting future trajectories of multiple agents,” in International Conference on Learning Representations, 2022

    work page 2022

  3. [3]

    Scept: Scene-consistent, policy- based trajectory predictions for planning,

    Y . Chen, B. Ivanovic, and M. Pavone, “Scept: Scene-consistent, policy- based trajectory predictions for planning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 17 103–17 112

    work page 2022

  4. [4]

    GPT-4 technical report,

    OpenAI, “GPT-4 technical report,” arXiv preprint arXiv:2303.08774 , 2023

    arXiv 2023

  5. [5]

    Gemini: a family of highly capable multimodal models,

    Google, “Gemini: a family of highly capable multimodal models,” arXiv preprint arXiv:2312.11805, 2023

    arXiv 2023

  6. [6]

    Training a helpful and harmless assistant with reinforcement learning from human feedback,

    Y . Bai, A. Jones, K. Ndousse, A. Askell, A. Chen, N. DasSarma, D. Drain, S. Fort, D. Ganguli, T. Henighan et al., “Training a helpful and harmless assistant with reinforcement learning from human feedback,” arXiv preprint arXiv:2204.05862 , 2022

    arXiv 2022

  7. [7]

    Training language models to follow instructions with human feedback,

    L. Ouyang, J. Wu, X. Jiang, D. Almeida, C. Wainwright, P. Mishkin, C. Zhang, S. Agarwal, K. Slama, A. Ray et al. , “Training language models to follow instructions with human feedback,” Advances in neural information processing systems , vol. 35, pp. 27 730–27 744, 2022

    2022

  8. [8]

    SimPO: Simple preference optimization with a reference-free reward,

    Y . Meng, M. Xia, and D. Chen, “SimPO: Simple preference optimization with a reference-free reward,” Advances in Neural Information Process- ing Systems, vol. 37, pp. 124 198–124 235, 2024

    work page 2024

Show all 34 references
  1. [9]

    Learning lane graph representations for motion forecasting,

    M. Liang, B. Yang, R. Hu, Y . Chen, R. Liao, S. Feng, and R. Urtasun, “Learning lane graph representations for motion forecasting,” in Com- puter Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part II 16 . Springer, 2020, pp. 541– 556

    2020

  2. [10]

    Motion transformer with global intention localization and local movement refinement,

    S. Shi, L. Jiang, D. Dai, and B. Schiele, “Motion transformer with global intention localization and local movement refinement,” Advances in Neural Information Processing Systems , vol. 35, pp. 6531–6543, 2022

    2022

  3. [11]

    Multipath: Multiple probabilistic anchor trajectory hypotheses for behavior prediction,

    Y . Chai, B. Sapp, M. Bansal, and D. Anguelov, “Multipath: Multiple probabilistic anchor trajectory hypotheses for behavior prediction,” in Conference on Robot Learning . PMLR, 2020, pp. 86–99

    2020

  4. [12]

    Multipath++: Efficient information fusion and trajectory aggregation for behavior prediction,

    B. Varadarajan, A. Hefny, A. Srivastava, K. S. Refaat, N. Nayakanti, A. Cornman, K. Chen, B. Douillard, C. P. Lam, D. Anguelov et al. , “Multipath++: Efficient information fusion and trajectory aggregation for behavior prediction,” in 2022 International Conference on Robotics ...

    2022

  5. [13]

    Wayformer: Motion forecasting via simple & efficient at- tention networks,

    N. Nayakanti, R. Al-Rfou, A. Zhou, K. Goel, K. S. Refaat, and B. Sapp, “Wayformer: Motion forecasting via simple & efficient at- tention networks,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2023, pp. 2980–2987

    2023

  6. [14]

    Query-centric trajectory prediction,

    Z. Zhou, J. Wang, Y .-H. Li, and Y .-K. Huang, “Query-centric trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 17 863–17 873

    2023

  7. [15]

    EDA: Evolving and distinct anchors for multimodal motion prediction,

    L. Lin, X. Lin, T. Lin, L. Huang, R. Xiong, and Y . Wang, “EDA: Evolving and distinct anchors for multimodal motion prediction,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 4, 2024, pp. 3432–3440

    2024

  8. [16]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, “End-to-end object detection with transformers,” in European conference on computer vision . Springer, 2020, pp. 213– 229

    2020

  9. [17]

    FJMP: Factorized joint multi-agent motion prediction over learned directed acyclic interac- tion graphs,

    L. Rowe, M. Ethier, E.-H. Dykhne, and K. Czarnecki, “FJMP: Factorized joint multi-agent motion prediction over learned directed acyclic interac- tion graphs,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 13 745–13 755

    2023

  10. [18]

    M2I: From factored marginal trajectory prediction to interactive prediction,

    Q. Sun, X. Huang, J. Gu, B. C. Williams, and H. Zhao, “M2I: From factored marginal trajectory prediction to interactive prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6543–6552

    2022

  11. [19]

    Qcnext: A next- generation framework for joint multi-agent trajectory prediction,

    Z. Zhou, Z. Wen, J. Wang, Y .-H. Li, and Y .-K. Huang, “Qcnext: A next- generation framework for joint multi-agent trajectory prediction,” arXiv preprint arXiv:2306.10508, 2023

    2023 arXiv

  12. [20]

    MTR++: Multi-agent motion prediction with symmetric scene modeling and guided intention query- ing,

    S. Shi, L. Jiang, D. Dai, and B. Schiele, “MTR++: Multi-agent motion prediction with symmetric scene modeling and guided intention query- ing,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 46, no. 5, pp. 3955–3971, 2024

    2024

  13. [21]

    Reasoning multi-agent behavioral topology for interactive autonomous driving,

    H. Liu, L. Chen, Y . Qiao, C. Lv, and H. Li, “Reasoning multi-agent behavioral topology for interactive autonomous driving,” in Advances in Neural Information Processing Systems , A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak, and C. Zhang, Eds., vol. 37....

    2024

  14. [22]

    INTERACTION Dataset: An INTERnational, Adversar- ial and Cooperative moTION Dataset in Interactive Driving Scenarios with Semantic Maps,

    W. Zhan, L. Sun, D. Wang, H. Shi, A. Clausse, M. Naumann, J. K ¨ummerle, H. K ¨onigshof, C. Stiller, A. de La Fortelle, and M. Tomizuka, “INTERACTION Dataset: An INTERnational, Adversar- ial and Cooperative moTION Dataset in Interactive Driving Scenarios with Semantic Maps,” a...

    1910 arXiv

  15. [23]

    Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset,

    S. Ettinger, S. Cheng, B. Caine, C. Liu, H. Zhao, S. Pradhan, Y . Chai, B. Sapp, C. R. Qi, Y . Zhou et al. , “Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset,” in Proceedings of the IEEE/CVF International Conference on Computer ...

    2021

  16. [24]

    Fine-tuning language models from human preferences,

    D. M. Ziegler, N. Stiennon, J. Wu, T. B. Brown, A. Radford, D. Amodei, P. Christiano, and G. Irving, “Fine-tuning language models from human preferences,” arXiv preprint arXiv:1909.08593 , 2019

    1909 arXiv

  17. [25]

    Rrhf: Rank responses to align language models with human feedback,

    H. Yuan, Z. Yuan, C. Tan, W. Wang, S. Huang, and F. Huang, “Rrhf: Rank responses to align language models with human feedback,” Advances in Neural Information Processing Systems , vol. 36, 2024

    2024

  18. [26]

    Model alignment as prospect theoretic optimization,

    K. Ethayarajh, W. Xu, N. Muennighoff, D. Jurafsky, and D. Kiela, “Model alignment as prospect theoretic optimization,” in Forty-first International Conference on Machine Learning , 2024

    2024

  19. [27]

    A general theoretical paradigm to understand learning from human preferences,

    M. G. Azar, Z. D. Guo, B. Piot, R. Munos, M. Rowland, M. Valko, and D. Calandriello, “A general theoretical paradigm to understand learning from human preferences,” in International Conference on Artificial Intelligence and Statistics . PMLR, 2024, pp. 4447–4455

    2024

  20. [28]

    Slic-hf: Sequence likelihood calibration with human feedback,

    Y . Zhao, R. Joshi, T. Liu, M. Khalman, M. Saleh, and P. J. Liu, “Slic-hf: Sequence likelihood calibration with human feedback,” arXiv preprint arXiv:2305.10425, 2023

    2023 arXiv

  21. [29]

    Direct preference optimization: Your language model is secretly a reward model,

    R. Rafailov, A. Sharma, E. Mitchell, C. D. Manning, S. Ermon, and C. Finn, “Direct preference optimization: Your language model is secretly a reward model,” Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  22. [30]

    Rank analysis of incomplete block designs: I. the method of paired comparisons,

    R. A. Bradley and M. E. Terry, “Rank analysis of incomplete block designs: I. the method of paired comparisons,” Biometrika, vol. 39, no. 3/4, pp. 324–345, 1952

    1952

  23. [31]

    The analysis of permutations,

    R. L. Plackett, “The analysis of permutations,” Journal of the Royal Statistical Society Series C: Applied Statistics , vol. 24, no. 2, pp. 193– 202, 1975

    1975

  24. [32]

    R. D. Luce, Individual choice behavior. Wiley New York, 1959, vol. 4

    1959

  25. [33]

    MotionDiffuser: Controllable multi-agent motion prediction using dif- fusion,

    C. Jiang, A. Cornman, C. Park, B. Sapp, Y . Zhou, D. Anguelov et al., “MotionDiffuser: Controllable multi-agent motion prediction using dif- fusion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 9644–9653

    2023

  26. [34]

    Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting,

    B. Wilson, W. Qi, T. Agarwal, J. Lambert, J. Singh, S. Khandelwal, B. Pan, R. Kumar, A. Hartnett, J. K. Pontes et al. , “Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting,” in Thirty-fifth Conference on Neural Information Processing Systems Data...

    2021

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.