Multi Task Agent Learning With Shared Trajectory Representation

via “multi-task learning with shared representations and task-specific heads”

PyTorch NLP framework with contextual embeddings.

Unique: Implements multi-task learning through a unified architecture where a shared BiLSTM encoder feeds into task-specific output heads (CRF for tagging, softmax for classification), enabling flexible combinations of different task types; supports dynamic task weighting during training to balance task contributions

vs others: More efficient than training separate models for each task while maintaining task-specific output constraints; enables knowledge transfer between related tasks, improving performance on low-resource tasks; simpler to implement than complex multi-task architectures with task-specific encoders

via “agent-training-loop orchestration and evaluation”

Library/framework for building language agents

Unique: Implements complete agent training loop mirroring neural network training with language-based gradients, enabling systematic improvement of agent behavior through experience on task distributions

vs others: More systematic than manual prompt iteration; more interpretable than RL-based agent training by preserving human-readable component updates

via “multi-task-learning-with-shared-representations”

A very simple framework for state-of-the-art NLP

Unique: Flair's multi-task learning framework uses shared embedding and encoder layers with task-specific output heads, enabling efficient knowledge transfer while maintaining task-specific prediction heads. This architecture allows fine-grained control over task weighting and loss functions, supporting both hard parameter sharing and soft parameter sharing strategies.

vs others: Flair's multi-task learning is more flexible than single-task pipelines (supports arbitrary task combinations) and more interpretable than end-to-end multi-task transformers, with explicit control over task weighting and loss functions.

via “multi-modal-trajectory-annotation-parsing”

Dataset by nvidia. 3,55,146 downloads.

Unique: Implements GR00T-X-specific annotation schema with native support for task hierarchies and robot morphology constraints, enabling semantic filtering of 334K trajectories without video I/O overhead — critical for large-scale embodied model training

vs others: Faster trajectory filtering than generic robotics datasets because annotations are pre-indexed and queryable without frame decompression, reducing data loading latency by 10-100x compared to frame-based filtering

via “multi-task visual policy learning with task-agnostic world models”

* ⏫ 02/2023: [Grounding Large Language Models in Interactive Environments with Online RL (GLAM)](https://arxiv.org/abs/2302.02662)

Unique: DreamerV3's task-agnostic world model learns shared visual representations without explicit task conditioning, relying on the policy learning objective to extract task-relevant information from the shared latent space. This contrasts with task-conditioned approaches (e.g., MTRL baselines) that explicitly encode task identity, making DreamerV3 more flexible for discovering emergent task structure.

vs others: Achieves better sample efficiency and generalization than task-conditioned baselines by learning task-invariant visual dynamics, while avoiding the computational overhead of task-specific world models or explicit task embeddings.

via “multi-task agent learning with shared trajectory representation”

### Other Papers <a name="2023op"></a>

Unique: Enables multi-task learning by conditioning the language model policy on task descriptions, allowing a single agent to learn from trajectories across diverse tasks and generalize to new tasks — this is distinct from task-specific agents that require separate training for each task

vs others: More sample-efficient than single-task agents because it leverages cross-task patterns, and more flexible than fixed multi-task architectures because task conditioning is learned end-to-end