Papers › Steering LLM Thinking with Budget Guidance

Steering LLM Thinking with Budget Guidance

16 Jun 2025arXiv:2506.13752archive 2025-07-28

Junyan Li, Wenshuo Zhao, Yang Zhang, Chuang Gan

Recent deep-thinking large language models often reason extensively to improve performance, but such lengthy reasoning is not always desirable, as it incurs excessive inference costs with disproportionate performance gains. Controlling reasoning length without sacrificing performance is therefore important, but remains challenging, especially under tight thinking budgets. We propose budget guidance, a simple yet effective method for steering the reasoning process of LLMs toward a target budget without requiring any LLM fine-tuning. Our approach introduces a lightweight predictor that models a Gamma distribution over the remaining thinking length during next-token generation. This signal is then used to guide generation in a soft, token-level manner, ensuring that the overall reasoning trace adheres to the specified thinking budget. Budget guidance enables natural control of the thinking length, along with significant token efficiency improvements over baseline methods on challenging math benchmarks. For instance, it achieves up to a 26% accuracy gain on the MATH-500 benchmark under tight budgets compared to baseline methods, while maintaining competitive accuracy with only 63% of the thinking tokens used by the full-thinking model. Budget guidance also generalizes to broader task domains and exhibits emergent capabilities, such as estimating question difficulty. The source code is available at: https://github.com/UMass-Embodied-AGI/BudgetGuidance.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2506.13752")

Code

Syntology Ran 11 of 16 code samples harvested from 1 repository linked to this paper; 5 have no recorded run. Of those that ran: 11 ran with no contract checked.

By repository: official repository: 16 samples from 1 repository, 11 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

umass-embodied-agi/budgetguidance officialmentioned in papermentioned on GitHubjaxMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

16 samples harvested; 11 ran; 0 honoured the contract we drafted; 5 have no recorded run. Read from Syntology's graph 2026-09-25; that is when this build read the record, not when the samples ran.

11ran
5unverified

Licence: 0 of the 16 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from umass-embodied-agi/budgetguidance. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

build_tip_message umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/deprecate_models.py official repository ran fingerprinted MIT (permissive) · ab306165fa74656c · report
flatten_dict umass-embodied-agi/budgetguidance/3rdparty/trl/trl/core.py official repository ran MIT (permissive) · dfd0ecd9cb25af69 · report
get_cased_name umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/modular_model_converter.py official repository ran fingerprinted MIT (permissive) · 8da0f9cdbba03509 · report
get_config_class_from_processor_class umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/create_dummy_models.py official repository ran MIT (permissive) · 3d999b95009c7537 · report
get_list_of_repo_model_paths umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/models_to_deprecate.py official repository ran fingerprinted MIT (permissive) · 938fc7c2422ef91d · report
get_model_doc_path umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/deprecate_models.py official repository ran MIT (permissive) · 051d790d9b197d9d · report
get_module_source_from_name umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/modular_model_converter.py official repository ran MIT (permissive) · 029be6c1f552852b · report
get_new_python_files_between_commits umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/pr_slow_ci_models.py official repository ran MIT (permissive) · 0f78c7037b35a137 · report
get_processor_types_from_config_class umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/create_dummy_models.py official repository ran MIT (permissive) · 0c0b81b081919d8e · report
masked_mean umass-embodied-agi/budgetguidance/3rdparty/trl/trl/core.py official repository ran fingerprinted MIT (permissive) · 4f7075aac2163cbc · report
masked_var umass-embodied-agi/budgetguidance/3rdparty/trl/trl/core.py official repository ran fingerprinted MIT (permissive) · 67607956afbb2a83 · report
extract_model_info umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/deprecate_models.py official repository unverified MIT (permissive) · 54b7273383fcfc2d · report
get_architectures_from_config_class umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/create_dummy_models.py official repository unverified MIT (permissive) · b176efb26f72e507 · report
has_think_block umass-embodied-agi/budgetguidance/training/run_data_augmentation.py official repository unverified MIT (permissive) · 544ec9956230c399 · report
modify_think_format umass-embodied-agi/budgetguidance/training/run_data_augmentation.py official repository unverified MIT (permissive) · 08837cd3c464a8e2 · report
preserve_case_replace umass-embodied-agi/budgetguidance/3rdparty/transformers/utils/modular_model_converter.py official repository unverified MIT (permissive) · d646d0a64800a8fe · report

Tasks

Math

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections