# AI capabilities can be significantly improved without expensive retraining

arXiv: https://arxiv.org/abs/2312.07413 · HTML: https://mlsanity.com/paper/2312.07413/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 48 references that are in the map.
- First posted on arXiv 12 Dec 2023.

## Builds on (most cited first, 30 of 33)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Tree of Thoughts: Deliberate Problem Solving with Large Language Models](https://mlsanity.com/paper/2305.10601/) (2023)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [A Survey on Large Language Model based Autonomous Agents](https://mlsanity.com/paper/2308.11432/) (2023)
- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [Solving Quantitative Reasoning Problems with Language Models](https://mlsanity.com/paper/2206.14858/) (2022)
- [Improving language models by retrieving from trillions of tokens](https://mlsanity.com/paper/2112.04426/) (2021)
- [Measuring and Narrowing the Compositionality Gap in Language Models](https://mlsanity.com/paper/2210.03350/) (2022)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [Show Your Work: Scratchpads for Intermediate Computation with Language Models](https://mlsanity.com/paper/2112.00114/) (2021)
- [Large Language Models Can Self-Improve](https://mlsanity.com/paper/2210.11610/) (2022)
- [Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models](https://mlsanity.com/paper/2310.04406/) (2023)
- [Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models](https://mlsanity.com/paper/2305.04091/) (2023)
- [Identifying the Risks of LM Agents with an LM-Emulated Sandbox](https://mlsanity.com/paper/2309.15817/) (2023)
- [Scaling Laws for Autoregressive Generative Modeling](https://mlsanity.com/paper/2010.14701/) (2020)
- [The False Promise of Imitating Proprietary LLMs](https://mlsanity.com/paper/2305.15717/) (2023)
- [InterCode: Standardizing and Benchmarking Interactive Coding with Execution Feedback](https://mlsanity.com/paper/2306.14898/) (2023)
- [Emergent autonomous scientific research capabilities of large language models](https://mlsanity.com/paper/2304.05332/) (2023)
- [Language Models Can Teach Themselves to Program Better](https://mlsanity.com/paper/2207.14502/) (2022)
- [Parsel: Algorithmic Reasoning with Language Models by Composing Decompositions](https://mlsanity.com/paper/2212.10561/) (2022)
- [The Gradient of Generative AI Release: Methods and Considerations](https://mlsanity.com/paper/2302.04844/) (2023)
- [Measuring the Algorithmic Efficiency of Neural Networks](https://mlsanity.com/paper/2005.04305/) (2020)
- [Scaling laws for single-agent reinforcement learning](https://mlsanity.com/paper/2301.13442/) (2023)

## Built on by (most cited first, 28 of 28)

- [Black-Box Access is Insufficient for Rigorous AI Audits](https://mlsanity.com/paper/2401.14446/) (2024)
- [Visibility into AI Agents](https://mlsanity.com/paper/2401.13138/) (2024)
- [Stress-Testing Capability Elicitation With Password-Locked Models](https://mlsanity.com/paper/2405.19550/) (2024)
- [Emergence of Hidden Capabilities: Exploring Learning Dynamics in Concept Space](https://mlsanity.com/paper/2406.19370/) (2024)
- [Towards evaluations-based safety cases for AI scheming](https://mlsanity.com/paper/2411.03336/) (2024)
- [Algorithmic progress in language models](https://mlsanity.com/paper/2403.05812/) (2024)
- [On the Limitations of Compute Thresholds as a Governance Strategy](https://mlsanity.com/paper/2407.05694/) (2024)
- [SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers](https://mlsanity.com/paper/2609.01343/) (2026)
- [Training Compute Thresholds: Features and Functions in AI Regulation](https://mlsanity.com/paper/2405.10799/) (2024)
- [Measuring AI agent autonomy: Towards a scalable approach with code inspection](https://mlsanity.com/paper/2502.15212/) (2025)
- [STREAM (ChemBio): A Standard for Transparently Reporting Evaluations in AI Model Reports](https://mlsanity.com/paper/2508.09853/) (2025)
- [Discovering Forbidden Topics in Language Models](https://mlsanity.com/paper/2505.17441/) (2025)
- [Measuring AI R&D Automation](https://mlsanity.com/paper/2603.03992/) (2026)
- [Hazards from Increasingly Accessible Fine-Tuning of Downloadable Foundation Models](https://mlsanity.com/paper/2312.14751/) (2023)
- [The Role of Governments in Increasing Interconnected Post-Deployment Monitoring of AI](https://mlsanity.com/paper/2410.04931/) (2024)
- [On the Origin of Algorithmic Progress in AI](https://mlsanity.com/paper/2511.21622/) (2025)
- [Declare and Justify: Explicit assumptions in AI evaluations are necessary for effective regulation](https://mlsanity.com/paper/2411.12820/) (2024)
- [What AI evaluations for preventing catastrophic risks can and cannot do](https://mlsanity.com/paper/2412.08653/) (2024)
- [Position: Ensuring mutual privacy is necessary for effective external evaluation of proprietary AI systems](https://mlsanity.com/paper/2503.01470/) (2025)
- [Internal Data Repetition Destroys Language Models](https://mlsanity.com/paper/2606.24998/) (2026)
- [Rethinking LLM Advancement: Compute-Dependent and Independent Paths to Progress](https://mlsanity.com/paper/2505.04075/) (2025)
- [Compute Requirements for Algorithmic Innovation in Frontier AI Models](https://mlsanity.com/paper/2507.10618/) (2025)
- [Multi-Head Attention Residuals](https://mlsanity.com/paper/2607.27230/) (2026)
- [Multilinguality at the Edge: Developing Language Models for the Global South](https://mlsanity.com/paper/2604.21637/) (2026)
- [Comprehensive AI governance requires addressing non-model gains](https://mlsanity.com/paper/2606.00047/) (2026)
- [The Dynamics of Intelligence Explosions](https://mlsanity.com/paper/2608.14426/) (2026)
- [What if automating AI R&D triggers an intelligence explosion?](https://mlsanity.com/paper/2609.36054/) (2026)
- [How Much Is an AI Token Worth? Scaling Laws for Wild AI-Generated Web Text](https://mlsanity.com/paper/2609.40295/) (2026)

## Abstract (arXiv)

State-of-the-art AI systems can be significantly improved without expensive retraining via "post-training enhancements"-techniques applied after initial training like fine-tuning the system to use a web browser. We review recent post-training enhancements, categorizing them into five types: tool-use, prompting methods, scaffolding, solution selection, and data generation. Different enhancements improve performance on different tasks, making it hard to compare their significance. So we translate improvements from different enhancements into a common currency, the compute-equivalent gain: how much additional training compute would be needed to improve performance by the same amount as the enhancement. Our non-experimental work shows that post-training enhancements have significant benefits: most surveyed enhancements improve benchmark performance by more than a 5x increase in training compute, some by more than 20x. Post-training enhancements are relatively cheap to develop: fine-tuning costs are typically <1% of the original training cost. Governing the development of capable post-training enhancements may be challenging because frontier models could be enhanced by a wide range of actors.