• About Us
  • Disclaimer
  • Contact Us
  • Privacy Policy
Friday, May 1, 2026
mGrowTech
No Result
View All Result
  • Technology And Software
    • Account Based Marketing
    • Channel Marketing
    • Marketing Automation
      • Al, Analytics and Automation
      • Ad Management
  • Digital Marketing
    • Social Media Management
    • Google Marketing
  • Direct Marketing
    • Brand Management
    • Marketing Attribution and Consulting
  • Mobile Marketing
  • Event Management
  • PR Solutions
  • Technology And Software
    • Account Based Marketing
    • Channel Marketing
    • Marketing Automation
      • Al, Analytics and Automation
      • Ad Management
  • Digital Marketing
    • Social Media Management
    • Google Marketing
  • Direct Marketing
    • Brand Management
    • Marketing Attribution and Consulting
  • Mobile Marketing
  • Event Management
  • PR Solutions
No Result
View All Result
mGrowTech
No Result
View All Result
Home Al, Analytics and Automation

Can We Improve Llama 3’s Reasoning Through Post-Training Alone? ASTRO Shows +16% to +20% Benchmark Gains

Josh by Josh
July 4, 2025
in Al, Analytics and Automation
0
Can We Improve Llama 3’s Reasoning Through Post-Training Alone? ASTRO Shows +16% to +20% Benchmark Gains


Improving the reasoning capabilities of large language models (LLMs) without architectural changes is a core challenge in advancing AI alignment and usability. Researchers at Meta AI and the University of Washington have introduced ASTRO—Autoregressive Search-Taught Reasoner—a novel post-training framework designed to enhance reasoning in Llama-3.1-70B-Instruct. ASTRO is unique in teaching models to perform in-context search, self-reflection, and backtracking, mechanisms often associated with human problem-solving and traditional symbolic search algorithms. Through this approach, ASTRO boosts Llama 3’s math performance on several competitive benchmarks with significant improvements:

  • MATH 500: 65.8% ➝ 81.8%
  • AMC 2023: 37.5% ➝ 64.4%
  • AIME 2024: 10.0% ➝ 30.0%

Search-Guided Chain-of-Thought Generation

ASTRO’s methodology begins with a Monte Carlo Tree Search (MCTS) over mathematical problem-solving trajectories. This search explores both correct and incorrect reasoning paths. The key innovation is procedure cloning: entire search trees are linearized into long chain-of-thoughts (CoT) that naturally encode both failures and recoveries via self-reflection and backtracking. These linearized traces are rewritten in natural language and used as the basis for supervised fine-tuning (SFT).

This results in a model that doesn’t just solve problems step-by-step but reevaluates its trajectory—often backtracking after self-assessment to correct intermediate reasoning mistakes. For instance, the model may interject with phrases like “Let’s go back to where we set up the equation” when its internal confidence drops.

Supervised Fine-Tuning: Injecting Search Priors

ASTRO fine-tunes Llama-3.1-70B-Instruct on 36.1K curated CoT solutions from MATH, AMC/AIME, and AoPS-style datasets. The model trained with ASTRO-SFT achieves:

  • MATH 500: 69.6%
  • AMC 2023: 51.9%
  • AIME 2024: 16.3%

These scores are competitive with or exceed those of baseline and SPOC/Step-KTO variants trained without explicit search priors. Importantly, even SFT alone—without reinforcement learning—yields performance boosts by exposing the model to search-structured reasoning data.

Reinforcement Learning with Search-Aware Initialization

ASTRO proceeds to reinforcement learning (RL) by initializing with the SFT checkpoint and running an RL loop using a modified Group Relative Policy Optimization (GRPO). Unlike standard preference-based RL, ASTRO employs verifiable reward signals (+1 for correct, -1 for incorrect) on 8.7K moderately difficult prompts. During training, the model’s CoT generation grows longer—from ~1.8K to ~6K tokens—demonstrating deeper internal exploration.

The resulting ASTRO-RL model achieves:

  • MATH 500: 81.8%
  • AMC 2023: 64.4%
  • AIME 2024: 30.0%

These results rival or exceed models with larger parameter counts and confirm the importance of ASTRO’s search-aware initialization.

Backtracking Behavior Correlates with Reasoning Success

A striking empirical observation is the positive correlation between backtracking frequency and performance. As training progresses, ASTRO-RL exhibits more self-corrective actions and deeper exploration. Pearson correlation coefficients across benchmarks exceed 0.8, indicating that self-reflection and backtracking are not merely cosmetic behaviors but functionally tied to better accuracy.

Comparative Insights and Broader Impact

Control experiments comparing ASTRO with models trained on direct CoT solutions (no search priors) reveal that even when trained on the same problem sets and search trees, ASTRO consistently outperforms. For instance, ASTRO-RL beats Direct-RL by:

  • +2% on MATH 500
  • +3.9% on AMC 2023
  • +2.9% on AIME 2024

Moreover, ASTRO’s outputs can be visualized as directed graphs, with nodes as reasoning steps and edges capturing transitions, reflections, and corrections—facilitating better interpretability.

ASTRO Key Takeaways Table

Conclusion

ASTRO demonstrates that LLMs like Llama 3 can learn to reason more effectively—not through larger models or longer pretraining, but via principled post-training techniques. By mimicking search algorithms in natural language, ASTRO enables models to think before answering, doubt their own steps, and correct themselves mid-reasoning. This framework sets a new benchmark for fine-tuning open LLMs to approach human-like reasoning through search-inspired behaviors.


Check out the Paper. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 100k+ ML SubReddit and Subscribe to our Newsletter.


Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over 2 million monthly views, illustrating its popularity among audiences.



Source_link

READ ALSO

Improving understanding with language | MIT News

A Coding Implementation on Pyright Type Checking Covering Generics, Protocols, Strict Mode, Type Narrowing, and Modern Python Typing

Related Posts

Improving understanding with language | MIT News
Al, Analytics and Automation

Improving understanding with language | MIT News

May 1, 2026
A Coding Implementation on Pyright Type Checking Covering Generics, Protocols, Strict Mode, Type Narrowing, and Modern Python Typing
Al, Analytics and Automation

A Coding Implementation on Pyright Type Checking Covering Generics, Protocols, Strict Mode, Type Narrowing, and Modern Python Typing

May 1, 2026
DeepSeek’s new AI model is rolling out quietly, not to the Wall Street market shock
Al, Analytics and Automation

DeepSeek’s new AI model is rolling out quietly, not to the Wall Street market shock

April 30, 2026
Solving the “Whac-a-mole dilemma”: A smarter way to debias AI vision models | MIT News
Al, Analytics and Automation

Solving the “Whac-a-mole dilemma”: A smarter way to debias AI vision models | MIT News

April 30, 2026
IBM Releases Two Granite Speech 4.1 2B Models: Autoregressive ASR with Translation and Non-Autoregressive Editing for Fast Inference
Al, Analytics and Automation

IBM Releases Two Granite Speech 4.1 2B Models: Autoregressive ASR with Translation and Non-Autoregressive Editing for Fast Inference

April 30, 2026
How AI Policy in South Africa Is Ruining Itself
Al, Analytics and Automation

How AI Policy in South Africa Is Ruining Itself

April 30, 2026
Next Post
Microsoft is closing its local operations in Pakistan

Microsoft is closing its local operations in Pakistan

POPULAR NEWS

Trump ends trade talks with Canada over a digital services tax

Trump ends trade talks with Canada over a digital services tax

June 28, 2025
Communication Effectiveness Skills For Business Leaders

Communication Effectiveness Skills For Business Leaders

June 10, 2025
15 Trending Songs on TikTok in 2025 (+ How to Use Them)

15 Trending Songs on TikTok in 2025 (+ How to Use Them)

June 18, 2025
App Development Cost in Singapore: Pricing Breakdown & Insights

App Development Cost in Singapore: Pricing Breakdown & Insights

June 22, 2025
Comparing the Top 7 Large Language Models LLMs/Systems for Coding in 2025

Comparing the Top 7 Large Language Models LLMs/Systems for Coding in 2025

November 4, 2025

EDITOR'S PICK

Mistral AI Unveils Mistral Medium 3.1: Enhancing AI with Superior Performance and Usability

Mistral AI Unveils Mistral Medium 3.1: Enhancing AI with Superior Performance and Usability

August 13, 2025
The Impact of Sleep Deprivation on Ghrelin and Leptin: A Look at Appetite Regulation

The Impact of Sleep Deprivation on Ghrelin and Leptin: A Look at Appetite Regulation

December 20, 2025
How Brands are Reaching Gen Z in 2025

How Brands are Reaching Gen Z in 2025

September 4, 2025
13 Local Citation Building Services You Can Trust

13 Local Citation Building Services You Can Trust

April 24, 2026

About

We bring you the best Premium WordPress Themes that perfect for news, magazine, personal blog, etc. Check our landing page for details.

Follow us

Categories

  • Account Based Marketing
  • Ad Management
  • Al, Analytics and Automation
  • Brand Management
  • Channel Marketing
  • Digital Marketing
  • Direct Marketing
  • Event Management
  • Google Marketing
  • Marketing Attribution and Consulting
  • Marketing Automation
  • Mobile Marketing
  • PR Solutions
  • Social Media Management
  • Technology And Software
  • Uncategorized

Recent Posts

  • ChatGPT Images 2.0 is a hit in India, but not a big winner elsewhere, yet
  • Improving understanding with language | MIT News
  • strategic rebrand resets Anmac’s place in Australia’s health system – Truly Deeply – Brand Strategy & Creative Agency Melbourne
  • Reitmans Unveils a New Logo and Enters a New Era With a Reimagined Store Concept
  • About Us
  • Disclaimer
  • Contact Us
  • Privacy Policy
No Result
View All Result
  • Technology And Software
    • Account Based Marketing
    • Channel Marketing
    • Marketing Automation
      • Al, Analytics and Automation
      • Ad Management
  • Digital Marketing
    • Social Media Management
    • Google Marketing
  • Direct Marketing
    • Brand Management
    • Marketing Attribution and Consulting
  • Mobile Marketing
  • Event Management
  • PR Solutions