MoP-JEPA: Hard-Assigned Predictor Mixtures for Stochastic JEPA World Models

• URL: http://arxiv.org/abs/2607.05238v1
• Date: Mon, 06 Jul 2026 15:47:37 GMT
• Title: MoP-JEPA: Hard-Assigned Predictor Mixtures for Stochastic JEPA World Models
• Abstract summary: JEPA world models predict the next latent state with a single deterministic predictor trained by latent regression.<n>We show that this fails structurally when the environment is branching.<n>We prove that MoP-JEPA's hard-assigned predictors converge instead to a quantizer of the transition distribution.
• Score: 12.290764474660103
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: JEPA world models predict the next latent state with a single deterministic predictor trained by latent regression. We show that this fails structurally when the environment is stochastic: at a branching transition, the regression-optimal predictor outputs the conditional mean of the successor embeddings, a point between the true next states that corresponds to no state at all. We prove this collapse for deterministic and gated mixture-of-experts predictors, and prove that MoP-JEPA's hard-assigned predictors converge instead to a quantizer of the transition distribution: one head per successor mode, enumerable in a single forward pass, which is the interface a planner consumes. On official OGBench offline data with leak-free evaluation, planning over single-predictor rollouts performs poorly ($0.02$--$0.09$ success) while planning over our predicted modes reaches up to $0.85$, ahead of deterministic, gated-MoE, and variational predictors on every task. Because multi-prediction evaluation invites coverage freeloading, a verification protocol is part of the method: an input-agnostic codebook control, a shuffled-context test, router-gated readouts, transition-precision guards, and a verified-route criterion in which the model proposes its transition graph blind and ground truth is used only to check the result. Under this criterion our method outperforms the strongest soft alternative on all three mazes ($2$--$5\times$), and the protocol identifies the remaining gap in that baseline's raw scores as routes through predicted transitions that do not exist. The same model executes in the real environment, placing second of seven against the published OGBench baselines on the hardest maze. Multimodal dynamics decide whether a JEPA world model can plan at all; a mixture of predictors with hard assignment is a minimal and verifiable fix.

Related papers

• Flow-JEPA: Flow Matching for Robust Latent Dynamics in JEPA World Models [18.035166689862205]
  Joint-Embedding Predictive Architectures (JEPAs) have shown strong potential for learning compact predictive representations.<n>We propose Flow-JEPA, a conditional flow matching dynamics model that jointly generates a sequence of future latent states conditioned on the current observation and actions.
  arXiv  Detail & Related papers  (2026-08-29T03:40:06Z)
• Qantara: Bridge-Flow Training for Multi-Paradigm JEPA Control [5.357451930622806]
  Joint-Embedding Predictive Architectures underpin a growing family of latent world models for control from raw pixels.<n>A single checkpoint that serves both would defer this choice to inference, when deployment constraints determine which path wins.<n>We present Qantara, an end-to-end JEPA whose joint training objective pairs a Brownian-bridge interpolant between consecutive clean latents.
  arXiv  Detail & Related papers  (2026-07-06T12:12:50Z)
• Predictive Objectives Discard Exogenous Control-Relevant Features: A Controlled Mechanistic Study [0.0]
  Joint-embedding predictive (JEPA) objectives learn representations by predicting future latents.<n>This occurs because the objective optimize temporal predictability rather than control-relevance.<n>We show that as little as 2% of reward-labeled transitions recovers the feature, the effect holds across two environments.
  arXiv  Detail & Related papers  (2026-06-29T10:00:50Z)
• Rethinking Training & Inference for Forecasting: Linking Winner-Take-All back to GMMs [48.245159309501666]
  Trajectory forecasting for autonomous driving has advanced rapidly, yet representative models often produce uninformative posteriors over forecast modes.<n>We argue that this K-means-like hard assignment (one-hot) is the source of uninformative mode probabilities.<n>We introduce two post-hoc treatments: (1) test-time posterior-weighted merging that aggregates nearby candidate trajectories; and (2) a one-step expectation-maximization (EM) update.
  arXiv  Detail & Related papers  (2026-06-24T22:26:43Z)
• PRISM: PRior-guided Imagination Sampling in world Models [7.304186330343712]
  We introduce PRISM, a task-agnostic framework that extracts both from a single dataset.<n>PRISM improves success rates by 35 percentage points over vanilla world-model-based MPC on Cube and 32 percentage points on PushT.
  arXiv  Detail & Related papers  (2026-06-06T04:18:10Z)
• UR-JEPA: Uniform Rectifiability as a Regularizer for Joint-Embedding Predictive Architectures [0.0]
  We propose emphUR-JEPA, which targets a uniformly $n$-rectifiable measure of local dimension tangent $n$ at small scales.<n>We show that on all four datasets UR--JEPA($mathcalLtextCGLT$) produces a global geometric spectrum with a $4$ to $5$ order-of-magnitude drop at index $sim 20$ to $25$ out of $D = 32$, while LeJEPA's spectrum is near-flat (top-to-bottom ratio at most $3.6$)
  arXiv  Detail & Related papers  (2026-05-31T20:26:54Z)
• Entropy Across the Bridge: Conditional-Marginal Discretization for Flow and Schrödinger Samplers [68.9946633641494]
  Flow matching and Schrdinger bridges define probability, yet their inference grids are usually or inherited from one-endpoint diffusion.<n>We derive a conditional-marginal entropy-rate objective for bridge-aware discretization, separating endpoint-conditioned bridge geometry from marginal flow evolution.<n>On EDM/CIFAR-10, the entropic time-discretization gives the best tested five-step FID (186.3 pm 4.0 versus 200.5 pm 2.9 for linear and 238.0 pm 5.3 for cosine)
  arXiv  Detail & Related papers  (2026-05-15T16:11:10Z)
• Representation Without Reward: A JEPA Audit for LLM Fine-Tuning [1.2691047660244335]
  Joint-embedding predictive architectures (JEPAs) propose that a model should learn more useful abstractions when trained to predict latent representations rather than observed outputs.<n>For autoregressive language-model fine-tuning the principle entails a stricter requirement: the induced hidden-state geometry must reach the language-model head emphand improve the decoded task metric.<n>We test that requirement under a fixed Llama-3.2-1B-Instruct LoRA harness on natural-language-to-regex generation.
  arXiv  Detail & Related papers  (2026-05-14T20:27:32Z)
• Active Multiple-Prediction-Powered Inference [2.1710524042938073]
  Post-deployment monitoring of healthcare requires statistically valid, label-efficient methods, but gold-standard labels from clinician chart review are expensive.<n>We propose Prediction Active-Prediction-Powered Inference (AM-PPI), which routes each instance to a cost-appropriate predictor, samples gold-standard labels in proportion, and repredicts predictions to a single deployment-time.
  arXiv  Detail & Related papers  (2026-05-08T19:48:57Z)
• Correction and Corruption: A Two-Rate View of Error Flow in LLM Protocols [51.56484100374058]
  We propose a paired-outcome measurement interface for auditing a single protocol step on exact-match tasks.<n>For each instance, the interface records a baseline correctness bit and a post-step correctness bit.<n>These rates predict accuracy changes and define a reusable empirical interface testable across seeds, mixtures, and pipelines.
  arXiv  Detail & Related papers  (2026-04-20T13:25:40Z)
• LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels [49.35636088613484]
  Joint Embedding Predictive Architectures (JEPAs) offer a compelling framework for learning world models in compact latent spaces.<n>We introduce LeModelWorld, the first JEPA that trains stably end-to-end from raw pixels using only two loss terms.<n>With 15M parameters trainable on a single GPU in a few hours, LeWM plans up to 48x faster than foundation-model-based world models.
  arXiv  Detail & Related papers  (2026-03-13T19:48:14Z)
• TABES: Trajectory-Aware Backward-on-Entropy Steering for Masked Diffusion Models [35.327100592206115]
  Backward-on-Entropy (BoE) Steering is a gradient-guided inference framework that approximates infinite-horizon context via a single backward pass.<n>To ensure scalability, we introduce ttexttActiveQueryAttention, a sparse adjoint primitive that exploits the structure of the masking objective to reduce backward pass complexity.
  arXiv  Detail & Related papers  (2026-01-30T19:10:32Z)
• Revisiting Weighted Strategy for Non-stationary Parametric Bandits and MDPs [56.246783503873225]
  This paper revisits the weighted strategy for non-stationary parametric bandits.<n>We propose a simpler weight-based algorithm that is as efficient as window/restart-based algorithms.<n>Our framework can be used to improve regret bounds of other parametric bandits.
  arXiv  Detail & Related papers  (2026-01-03T04:50:21Z)
• Agentic World Modeling for 6G: Near-Real-Time Generative State-Space Reasoning [70.56067503630486]
  We argue that sixth-generation (6G) intelligence is not fluent token prediction but calibrated the capacity to imagine and choose.<n>We show that WM-MS3M cuts mean absolute error (MAE) by 1.69% versus MS3M with 32% fewer parameters and similar latency, and achieves 35-80% lower root mean squared error (RMSE) than attention/hybrid baselines with 2.3-4.1x faster inference.
  arXiv  Detail & Related papers  (2025-11-04T17:22:22Z)
• Gaussian Embeddings: How JEPAs Secretly Learn Your Data Density [51.15085346971361]
  Joint Embedding Predictive Architectures (JEPAs) learn representations able to solve numerous downstream tasks out-of-the-box.<n>JEPAs combine two objectives: (i) a latent-space prediction term, i.e., the representation of a slightly perturbed sample must be predictable from the original sample's representation, and (ii) an anti-collapse term, i.e., not all samples should have the same representation.
  arXiv  Detail & Related papers  (2025-10-07T14:06:30Z)
• One Sample is Enough to Make Conformal Prediction Robust [53.78604391939934]
  We show that conformal prediction attains some robustness even with a forward pass on a single randomly perturbed input.<n>Our approach returns robust sets with smaller average set size compared to SOTA methods which use many (e.g. around 100) passes per input.
  arXiv  Detail & Related papers  (2025-06-19T19:14:25Z)
• Rejection via Learning Density Ratios [50.91522897152437]
  Classification with rejection emerges as a learning paradigm which allows models to abstain from making predictions.<n>We propose a different distributional perspective, where we seek to find an idealized data distribution which maximizes a pretrained model's performance.<n>Our framework is tested empirically over clean and noisy datasets.
  arXiv  Detail & Related papers  (2024-05-29T01:32:17Z)
• Joint Metrics Matter: A Better Standard for Trajectory Forecasting [67.1375677218281]
  Multi-modal trajectory forecasting methods evaluate using single-agent metrics (marginal metrics) Only focusing on marginal metrics can lead to unnatural predictions, such as colliding trajectories or diverging trajectories for people who are clearly walking together as a group. We present the first comprehensive evaluation of state-of-the-art trajectory forecasting methods with respect to multi-agent metrics (joint metrics): JADE, JFDE, and collision rate.
  arXiv  Detail & Related papers  (2023-05-10T16:27:55Z)
• Collaborative Uncertainty Benefits Multi-Agent Multi-Modal Trajectory Forecasting [61.02295959343446]
  This work first proposes a novel concept, collaborative uncertainty (CU), which models the uncertainty resulting from interaction modules.<n>We build a general CU-aware regression framework with an original permutation-equivariant uncertainty estimator to do both tasks of regression and uncertainty estimation.<n>We apply the proposed framework to current SOTA multi-agent trajectory forecasting systems as a plugin module.
  arXiv  Detail & Related papers  (2022-07-11T21:17:41Z)

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.