SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.18319 · 2024

Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Search

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 13 functions out of this paper's own repositories and ran 11 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
phunterlau/paper_without_code pwc_unofficial 11 of 13
FunctionStatusWhere it lives
analyze_execution_results Ran phunterlau/paper_without_code/paperwocode/code_executor.py
code served (permissive licence) · get_code("d878e7e7996abb1c")
benchmark_algorithm Ran phunterlau/paper_without_code/templates/algorithm.py
code served (permissive licence) · get_code("26a082825804b823")
create_conda_environment Ran phunterlau/paper_without_code/paperwocode/code_executor.py
code served (permissive licence) · get_code("123ddb8ddb4564c3")
download_pdf Ran phunterlau/paper_without_code/paperwocode/pdf_downloader.py
code served (permissive licence) · get_code("fcc6f5537839fe58")
evaluate_model Ran phunterlau/paper_without_code/output/workflow/2411.16905/generated_code.py
code served (permissive licence) · get_code("b20fc8d9642b452c")
extract_arxiv_id Ran phunterlau/paper_without_code/paperwocode/pdf_downloader.py
code served (permissive licence) · get_code("2332c3f0e7d84ce3")
generate_random_array Ran phunterlau/paper_without_code/templates/algorithm.py
code served (permissive licence) · get_code("5a2d755aeb05fd00")
generate_sample_data Ran phunterlau/paper_without_code/output/workflow/2411.16905/generated_code.py
code served (permissive licence) · get_code("df4a5c0a256b2e14")
run_code_with_timeout Ran phunterlau/paper_without_code/paperwocode/code_executor.py
code served (permissive licence) · get_code("1f13628b7e50b55a")
setup_logger Ran phunterlau/paper_without_code/paperwocode/code_generator.py
code served (permissive licence) · get_code("69d1f8af6cf4447f")
train_model Ran phunterlau/paper_without_code/output/workflow/2411.16905/generated_code.py
code served (permissive licence) · get_code("8fc873cde1192ad2")
estimate_api_cost Not yet run phunterlau/paper_without_code/paperwocode/code_generator.py
code served (permissive licence) · get_code("bc9d69072ce82a1f")
measure_execution_time Not yet run phunterlau/paper_without_code/templates/algorithm.py
code served (permissive licence) · get_code("2195c959742abc0f")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

In this work, we aim to develop an MLLM that understands and solves questions by learning to create each intermediate step of the reasoning involved till the final answer. To this end, we propose Collective Monte Carlo Tree Search (CoMCTS), a new learning-to-reason method for MLLMs, which introduces the concept of collective learning into ``tree search'' for effective and efficient reasoning-path searching and learning. The core idea of CoMCTS is to leverage collective knowledge from multiple models to collaboratively conjecture, search and identify effective reasoning paths toward correct answers via four iterative operations including Expansion, Simulation and Error Positioning, Backpropagation, and Selection. Using CoMCTS, we construct Mulberry-260k, a multimodal dataset with a tree of rich, explicit and well-defined reasoning nodes for each question. With Mulberry-260k, we perform collective SFT to train our model, Mulberry, a series of MLLMs with o1-like step-by-step Reasoning and Reflection capabilities. Extensive experiments demonstrate the superiority of our proposed methods on various benchmarks. Code will be available at https://github.com/HJYao00/Mulberry

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2412.18319")
get_code_for_paper("2412.18319")
have("2412.18319")

Connect an agent — have() is free.