101_HierarchicalReasoning_vs_CoT.txt

[[{PM.TODO]]
## 101: New AI architecture delivers 100x faster reasoning than LLMs with just 1,000 training examples

* <https://venturebeat.com/ai/new-ai-architecture-delivers-100x-faster-reasoning-than-llms-with-just-1000-training-examples/>

To test their model, researchers used benchmarks requiring extensive search 
and backtracking, such as Abstraction and Reasoning Corpus (ARC-AGI), extremely
difficult Sudoku puzzles and complex maze-solving tasks.

**The results show that HRM learns to solve problems that are intractable for 
even advanced LLMs**.
 For instance, on the “Sudoku-Extreme” and “Maze-Hard” benchmarks, 
state-of-the-art CoT models failed completely, scoring 0% accuracy.
 In contrast, HRM achieved near-perfect accuracy after being trained on
just 1,000 examples for each task.

On the ARC-AGI benchmark, a test of abstract reasoning and generalization, the 
27M-parameter HRM scored 40.3%. This surpasses leading CoT-based models like 
the much larger o3-mini-high (34.5%) and Claude 3.7 Sonnet (21.2%). This 
performance, achieved without a large pre-training corpus and with very limited 
data, highlights the power and efficiency of its architecture.



New AI architecture delivers 100x faster reasoning than LLMs with just 1,000 
training examples

... The architecture, known as the Hierarchical Reasoning Model (HRM), is inspired 
by how the human brain utilizes distinct systems for slow, deliberate planning 
and fast, intuitive computation. 

chain-of-thought reasoning, breaking down problems into intermediate text-based steps,
essentially forcing the model to “think out loud” as it works toward a solution ...
has fundamental limitations ... researchers at Sapient Intelligence argue that 
**CoT for reasoning is a crutch, not a satisfactory solution. It relies on 
brittle, human-defined decompositions where a single misstep or a misorder of 
the steps can derail the reasoning process entirely**. ... 
This dependency on generating explicit language tethers the model’s reasoning 
to the token level, often requiring massive amounts of training data and 
producing long, slow responses. This approach also overlooks the type of 
“latent reasoning” that occurs internally, without being explicitly 
articulated in language.

To move beyond CoT, the researchers explored “latent reasoning,” where 
instead of generating “thinking tokens,” the model reasons in its internal, 
abstract representation of the problem ... more aligned with how humans 
think; as the paper states, “**the brain sustains lengthy, coherent chains of 
reasoning with remarkable efficiency in a latent space, without constant 
translation back to language**.”
..human brain organizes computation hierarchically across cortical 
regions **operating at different timescales**, enabling deep, multi-stage 
reasoning.

HRM designed two coupled, recurrent modules: a high-level (H) module for
slow, abstract planning, and a low-level (L) module for fast, detailed
computations enabling a process the team calls “hierarchical convergence.”
... the slow H-module takes this result, updates its overall strategy,
and gives the L-module a new, refined sub-problem to work on.
** his effectively resets the L-module, preventing it from getting stuck 
(early convergence)** and allowing the entire system to perform a long
sequence of reasoning steps with a lean model architecture that doesn’t suffer from 
vanishing gradients.
HRM (left) smoothly converges on the solution across computation cycles and 
avoids early convergence (center, RNNs) and vanishing gradients (right, classic 
deep neural networks) Source: arXiv


A natural question is whether this “latent reasoning” comes at the cost of 
interpretability.

HRM outperforms large models on complex reasoning tasks Source: arXiv

While solving puzzles demonstrates the model’s power, the real-world 
implications lie in a different class of problems. According to Wang, 
developers should continue using LLMs for language-based or creative tasks, but 
for “complex or deterministic tasks,” an HRM-like architecture offers 
superior performance with fewer hallucinations. He points to “sequential 
problems requiring complex decision-making or long-term planning,” especially 
in latency-sensitive fields like embodied AI and robotics, or data-scarce 
domains like scientific exploration.

In these scenarios, HRM doesn’t just solve problems; it learns to solve them 
better. “In our Sudoku experiments at the master level… HRM needs 
progressively fewer steps as training advances—akin to a novice becoming an 
expert,” Wang explained.

For the enterprise, this is where the architecture’s efficiency translates 
directly to the bottom line. Instead of the serial, token-by-token generation 
of CoT, HRM’s parallel processing allows for what Wang estimates could be a 
“100x speedup in task completion time.” This means lower inference latency 
and the ability to run powerful reasoning on edge devices.

The cost savings are also substantial. “Specialized reasoning engines such as 
HRM offer a more promising alternative for specific complex reasoning tasks 
compared to large, costly, and latency-intensive API-based models,” Wang 
said. To put the efficiency into perspective, he noted that training the model 
for professional-level Sudoku takes roughly two GPU hours, and for the complex 
ARC-AGI benchmark, between 50 and 200 GPU hours—a fraction of the resources 
needed for massive foundation models. This opens a path to solving specialized 
business problems, from logistics optimization to complex system diagnostics, 
where both data and budget are finite.

Looking ahead, Sapient Intelligence is already working to evolve HRM from a 
specialized problem-solver into a more general-purpose reasoning module. “We 
are actively developing brain-inspired models built upon HRM,” Wang said, 
highlighting promising initial results in healthcare, climate forecasting, and 
robotics. He teased that these next-generation models will differ significantly 
from today’s text-based systems, notably through the inclusion of 
self-correcting capabilities.

The work suggests that for a class of problems that have stumped today’s AI 
giants, the path forward may not be bigger models, but smarter, more structured 
architectures inspired by the ultimate reasoning engine: the human brain.

Daily insights on business use cases with VB Daily

If you want to impress your boss, VB Daily has you covered. We give you the 
inside scoop on what companies are doing with generative AI, from regulatory 
shifts to practical deployments, so you can share insights for maximum ROI.

Read our Privacy Policy

Press Releases
Contact Us
Advertise
Share a News Tip
Contribute to DataDecisionMakers

Privacy Policy
Terms of Service

© 2025 VentureBeat. All rights reserved.

## Summarizing PDFs with Ollama and LangChain

* <https://vincent.codes.finance/posts/documents-llm/#loading-pdfs>





[[PM.TODO}]]