REVIEW 34 cited by
MiMo: Unlocking the Reasoning Potential of Language Model -- From Pretraining to Posttraining
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
MiMo: Unlocking the Reasoning Potential of Language Model -- From Pretraining to Posttraining
read the original abstract
We present MiMo-7B, a large language model born for reasoning tasks, with optimization across both pre-training and post-training stages. During pre-training, we enhance the data preprocessing pipeline and employ a three-stage data mixing strategy to strengthen the base model's reasoning potential. MiMo-7B-Base is pre-trained on 25 trillion tokens, with additional Multi-Token Prediction objective for enhanced performance and accelerated inference speed. During post-training, we curate a dataset of 130K verifiable mathematics and programming problems for reinforcement learning, integrating a test-difficulty-driven code-reward scheme to alleviate sparse-reward issues and employing strategic data resampling to stabilize training. Extensive evaluations show that MiMo-7B-Base possesses exceptional reasoning potential, outperforming even much larger 32B models. The final RL-tuned model, MiMo-7B-RL, achieves superior performance on mathematics, code and general reasoning tasks, surpassing the performance of OpenAI o1-mini. The model checkpoints are available at https://github.com/xiaomimimo/MiMo.
Forward citations
Cited by 34 Pith papers
-
DreamReasoner-8B: Block-Size Curriculum Learning for Diffusion Reasoning Models
Block-size curriculum learning trains an 8B diffusion model to achieve competitive reasoning performance on math and code benchmarks by transitioning from small to large training block sizes.
-
LLMZero: Discovering Adaptive Training Strategies for RL Post-Training via LLM Agents
LLMZero uses LLM agents to search training trajectories and discovers that capacity parameters accumulate monotonically while regularization parameters oscillate, leading to performance improvements of 9-140% on GRPO tasks.
-
Beyond Absolute Imitation: Anchored Residual Guidance for Privileged On-Policy Distillation
AR-OPD disentangles privileged supervision via anchored residual guidance to reduce hindsight leakage in on-policy distillation, reporting gains of 2.3 points over full privileged OPD and 7.9 over SFT on reasoning tasks.
-
SciVQR: A Multidisciplinary Multimodal Benchmark for Advanced Scientific Reasoning Evaluation
SciVQR is a new benchmark dataset for evaluating multimodal AI models on complex scientific reasoning tasks across six disciplines, including expert solutions for nearly half the items.
-
TRACER: Verifiable Generative Provenance for Multimodal Tool-Using Agents
TRACER attaches verifiable sentence-level provenance records to multimodal agent outputs using tool-turn alignment and semantic relations, yielding 78.23% answer accuracy and fewer tool calls than baselines on TRACE-Bench.
-
MMTutorBench: The First Multimodal Benchmark for AI Math Tutoring
MMTutorBench is the first multimodal benchmark for AI math tutoring with 685 problems, problem-specific rubrics across six dimensions, and evaluations of 12 MLLMs revealing performance gaps versus humans.
-
TennisTV: Do Multimodal Large Language Models Understand Tennis Rallies?
Introduces TennisTV benchmark for evaluating 17 MLLMs on tennis video understanding from stroke-level to rally-level tasks with automated pipelines and human verification.
-
Behavioural Signatures of Risk-Sensitive Decision-Making in Large Language Models
LLMs display stable Participation and Proactiveness risk profiles in multi-agent poker that remain largely robust to opponent mix and adapt heterogeneously under global and personal risk pressure.
-
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
OralAgent, a ReAct-style dental agent with 22 vision tools and a 134.8M-token textbook RAG corpus, reaches SOTA on MMOral-Uni, MMOral-OPG, and OralQA-ZH.
-
CodePercept: Code-Grounded Visual STEM Perception for MLLMs
Perception, not reasoning, is the main bottleneck for MLLM STEM visual reasoning, and training on executable reconstruction code measurably fixes it.
-
PIVOT: Efficient Query-Group Indexing for Token-Level Sparse Attention
Sharing one proxy full-prefix scan across nearby query groups matches dense DSA indexer accuracy while accelerating indexing up to 4× and end-to-end latency up to 1.6×.
-
When Preferences Fail to Become Incentives: A Utility-Behavior Gap in Large Language Models
Elicited preferences in LLMs do not function as effective incentives for higher-quality outputs on realistic writing tasks.
-
ELVA: Exploring Ranking-Driven Universal Multimodal Retrieval
ELVA applies ranking-driven RLVR to multimodal retrieval to reduce grain blindness in contrastive learning, reporting SOTA results and a 13.1% gain on the new MRBench benchmark.
-
Vortex: Efficient and Programmable Sparse Attention Serving for AI Agents
Vortex provides a programmable frontend and backend for sparse attention in LLM serving, delivering up to 3.46x throughput over full attention while preserving accuracy.
-
ADWIN: Adaptive Windows for Horizon-Aware On-Policy Distillation
ADWIN adaptively selects training horizons in on-policy distillation via prefix alignment checks, cutting end-to-end cost by up to 4.1x while matching or exceeding full-rollout accuracy on math and code benchmarks.
-
MACReD: A Multi-Agent Collaborative Reasoning Framework for Reaction Diagram Parsing
MACReD is a multi-agent collaborative reasoning framework for reaction diagram parsing that reports state-of-the-art F1 scores of 75.2% and 84.6% on the RxnScribe benchmark.
-
ROSE: Rollout On Serving GPUs via Cooperative Elasticity for Agentic RL
ROSE is a system for cooperative elasticity that co-locates serving and rollout models on shared GPUs, delivering 1.3-3.3x higher end-to-end throughput than fixed-resource baselines while preserving serving SLOs.
-
ROSE: Rollout On Serving GPUs via Cooperative Elasticity for Agentic RL
ROSE delivers 1.2-3.3x higher end-to-end throughput for agentic RL by safely co-using underutilized serving GPUs for rollouts while meeting serving SLOs.
-
CC-OCR V2: Benchmarking Large Multimodal Models for Literacy in Real-world Document Processing
CC-OCR V2 reveals that state-of-the-art large multimodal models substantially underperform on challenging real-world document processing tasks.
-
LayerBoost: Layer-Aware Attention Reduction for Efficient LLMs
LayerBoost applies layer-specific attention changes guided by sensitivity analysis plus brief distillation to cut LLM inference latency up to 68% while keeping competitive quality.
-
RefBench-PRO: Perceptual and Reasoning Oriented Benchmark for Referring Expression Comprehension
RefBench-PRO organizes REC into attribute, position, interaction, relation, commonsense, and reject tasks; no tested MLLM exceeds 72%, and Ref-R1 raises Qwen2.5-VL-7B from 57.6 to 69.4 on it.
-
X$^3$-OPD: Distilling Reasoning into Large Audio-Language Models via On-Policy Alignment
X3-OPD improves audio-grounded reasoning by training the audio student on its own rollouts with token-level teacher feedback, using a three-tier paired text-audio corpus.
-
Self-Improving is Often Sudden: Enlightenment-style Finetuning for Large-Scale Models
A training-free intervention that scales VLM residual connections and mixes LLM attention heads claims 1–3 point zero-shot accuracy gains, but the evidence is partly tuned on the reported benchmarks.
-
Enhancing Large Multimodal Models in Key Information Extraction via Scene-Aware Document Synthesis
Scene-aware multi-agent document synthesis plus error-driven hard-example expansion improves compact Qwen3-VL models on constrained and open-category KIE, topping reported on-device baselines.
-
ELVA: Exploring Ranking-Driven Universal Multimodal Retrieval
ELVA uses rule-based RL rewards to rank negatives by similarity, reducing grain blindness in universal multimodal retrieval and reporting a 13.1% gain on a new multi-grain benchmark.
-
YouZhi: Towards High-Concurrency Financial LLMs via Adaptive GQA-to-MLA Transition
YouZhi-LLM applies a layer-adaptive GQA-to-MLA transition plus Ascend-specific distillation and fine-tuning to reduce KV-cache size, yielding up to 2.69× higher concurrency and modest gains on financial benchmarks ver...
-
Treat Traffic Like Trees: A Semantic-Preserving Hierarchical Graph-Based Expert Framework for Encrypted Traffic Analysis
PTGAMoE is a hierarchical graph-based expert model that preserves protocol semantics for encrypted traffic classification and outperforms prior methods on benchmarks.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
-
I-WebGenBench : Evaluating Interactivity in LLM-Generated Scientific Web Applications
A Paper-to-Interactive-System Agent and I-WebGenBench benchmark with 19 papers enable converting scientific PDFs into executable interactive web systems, with PaperVoyager framework shown to improve quality.
-
SciVQR: A Multidisciplinary Multimodal Benchmark for Advanced Scientific Reasoning Evaluation
SciVQR is a new multimodal benchmark covering 54 scientific subfields that evaluates MLLMs on visual comprehension and multi-step reasoning, revealing significant limitations in leading models.
-
LayerBoost: Layer-Aware Attention Reduction for Efficient LLMs
LayerBoost selectively replaces or removes attention in non-critical transformer layers to cut inference latency up to 68% while recovering quality via brief distillation.
-
MiMo-V2-Flash Technical Report
MiMo-V2-Flash is a 309B/15B MoE model trained on 27T tokens with hybrid attention and multi-teacher on-policy distillation that matches larger models like DeepSeek-V3.2 while enabling 2.6x faster decoding via repurpos...
-
DSBench: A Comprehensive Benchmark for Evaluating External and In-Cabin Risks
A new benchmark claims to be the first to test VLMs on both external and in-cabin driving risks, and reports a fine-tuned model far outperforming all baselines.
-
Reinforcement Learning from Human Feedback
The book introduces the origins, mathematical setup, and optimization stages of RLHF including reward modeling, reinforcement learning, rejection sampling, and direct alignment algorithms.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.