Provides 130,000 typed RAG claim-verification decisions that pair source-derived reference labels with jev-1.13.0 model predictions. Includes sentence/span grounding, hallucination-type presence, whole-response grounding and response-quality decisions with retained provenance. Suited for grounding error analysis, prototype evidence-checkers and confidence-threshold studies; license: other.
Teaches small reasoning models to diagnose when additional internal thinking is insufficient and to selectively query stronger models; introduces FlyBy, a supervised + cost-aware RL framework that learns when/what to ask. Improves pass@ metrics on hard benchmarks while reducing serving cost.
Groups visually grounded appearances of the same physical instance into persistent, retrievable “biographies” so agents can follow objects across hours or days for long-video question answering. Links identity-aware observations to episodic context and visual evidence; improves EgoLifeQA to 72.0% and increases evidence-window reach from 37.6% to 58.9%.
Predicts compact 'prospective tokens' that summarize upcoming information needs and uses them to select a small set of past frames for conditioning long-horizon video generation, improving long-range consistency, visual quality, and action alignment while remaining plug-and-play across diverse generators.
Analyzes how proposer–solver loops in self-evolving search agents can develop shared errors (co-cheating) that inflate internal rewards; introduces Multi-Sample Verification and CrossFit (cross-fitted scoring with partitioned sources) to reduce false agreement and improve downstream search performance.
Aggregated, screened corpus of 55,050 normalized Indian public-information text bodies and 65,209 source records for retrieval and question-answering. Exports include deduplicated CSV/Parquet with provenance, topic labels, extraction quality flags and a private SQLite backup.
Adaptive-granularity post-retrieval evidence compressor for multimodal RAG that selects variable-sized regions from retrieved text, tables, images, and videos to retain necessary context while cutting irrelevant content. Key features include a hierarchy-based node encoder with parent-relative refinement and a critic that issues targeted follow-up retrievals; yields higher QA accuracy across five benchmarks and reduces reader-input tokens by ~14–28%.
Processes English and German text with long-context reasoning and structured tool-calling. Uses a 78B mixture-of-experts architecture that activates ~3.46B parameters per token, offers native 262k-token context (validated to 1M), and is released as Apache-2.0 weights — suited for RAG, document processing and human-in-the-loop decision support.
Analyzes how LLM agents prefer items from particular sources during end-to-end search and how these preferences affect selections across shopping, accommodation, and scholarly domains. Shows source labels can override item quality and evaluates mitigation strategies such as hiding sources, relabeling, supplying missing information, and counter-prompts.
GGUF-packaged weights for EmbeddingGemma 2 enabling local multimodal embeddings (text, image, video, audio); supports 768/512/256/128 dimensions, BF16/FP32 inference, selective modality loading for lower memory, and is suited for semantic search and RAG.
Enables multi-head, depth-wise residual retrieval in Transformers by expanding a single residual stream into multiple context-aware streams using causal depthwise convolutions. Improves pretraining convergence and downstream benchmark scores with minimal overhead, yielding notable iso-loss compute speedups in large MoE models.
Compact long-term-memory writer that decides what to retain and how to update persistent memory for external-memory frameworks. Trained with on-policy distillation and attributed reinforcement learning to improve cross-framework transfer (Mem0, MemOS, Hy-Memory, EverMemOS). Suited as a memory-writer component inside retrieval-augmented systems.