SYNTHESIS NOTE
Do large language models actually perform iterative optimization?
Explores whether LLMs execute genuine numerical procedures like Newton-Raphson or instead pattern-match to memorized solution templates when solving constrained optimization problems.
The constraint-optimization study identifies the mechanism behind the 55-60% plateau directly. LLMs cannot actually perform Newton-Raphson iterations in their latent space. They cannot execute primal-dual updates, nor any other iterative numerical procedure that genuine optimization requires. When asked to do so, they fall back to what the paper calls "result guessing" — recognizing the problem as similar to a standard power grid (or financial dataset, or security scenario) and emitting values that pattern-match what a valid solution should look like.
The fallback is silent. The output is fluent, well-formatted, often plausible. It can pass surface-level inspection because the model has seen many examples of what answers in this domain look like. What it has not done is solve the problem. The constraint values are wrong in ways that physical or financial systems would actually reject.
This explains why scale, architecture, and training regime do not move the plateau. They improve the template but not the procedure. A larger model has seen more example solutions and can produce more convincing guesses. Reinforcement learning on outcome rewards reinforces the template-matching pattern. None of this installs the iterative-computation capability the problem requires.
The mechanism — pattern-match against memorized solution-shapes when genuine computation is required — generalizes beyond optimization. It is plausibly the same mechanism behind a class of mathematical-reasoning failures where models produce confidently wrong numerical answers that resemble the right shape. The category is "looks like a solution; is not derived from one."
Inquiring lines that read this note 146
This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.
Is model misalignment inherent to base models or induced by alignment procedures?- Can communication problems and optimization problems be addressed with the same alignment approaches?
- What distinguishes minimal-pair asymmetry from standard accuracy evaluation?
- Can closed-form solutions compete with gradient descent optimization?
- Why do token-level language models fail at utterance-level pragmatic optimization?
- Why do language models fall back on frequency heuristics under structural complexity?
- Can instance seeds work for tasks beyond language understanding benchmarks?
- Do LLMs rely on surface heuristics instead of learning recursive grammar rules?
- Do instruction-tuned models learn tasks or just output format distributions?
- Can textual gradients generalize natural language feedback across computation graphs?
- Is gradient behavior in language functional or a sign of ambiguity?
- How should skill libraries coordinate with gradient-based weight optimization?
- Why does teacher forcing fail to capture long-range dependencies?
- Can language models execute iterative numerical methods in latent space?
- Does instruction tuning optimize language models for rhetorical polish over logical consistency?
- How do unstated constraints become invisible to training data distributions?
- Can benchmarks designed for shortcut learning detect heuristic override failures?
- Can explicit constraint statements override the dominance of surface heuristics?
- Why do task-specific heuristics fail at generalizing to sparse data regions?
- Do task-specific heuristics emerge because they compress well enough?
- What distinguishes intrinsic search from extrinsic search method approaches?
- Why do sigmoid conflict curves look the same across different language models?
- How do description-based identifiers bias language model output distribution?
- Why do smaller models favor code formats while larger models prefer natural language?
- Why do language models fail at iterative numerical optimization despite scale?
- Do pretrained language models carry reusable computational scaffolding for length handling?
- Are newer larger language models actually worse at faithful summarization?
- How do language models infer a benchmark's purpose without seeing its examples?
- How should benchmarks test whether models fit algorithms or patterns?
- Does the Heuristic Override Benchmark measure enumeration or world knowledge?
- How do general language model benchmarks predict specialized domain performance?
- Do smaller models optimized for benchmarks perform worse than their scores suggest in real tasks?
- Can universal function approximators be expensive to learn in practice?
- What non-parametric methods could replace latent factors for inductive learning?
- Why do hybrid memory and compute sparsity outperform pure parameter scaling?
- How do LLM activations sparsify differently under out-of-distribution inputs?
- Can a trained decoder replace both search and parameter updates?
- How does tournament selection without gradients compare to gradient-based hyperparameter tuning?
- Why do LLM outputs match researcher priors without solving tasks correctly?
- Why do large language models still have systematic blind spots with complex structures?
- Do standard language benchmarks underestimate what LLMs can actually do?
- Why do standard NLP benchmarks hide the most critical language limitations?
- What happens when we treat LLM outputs as sampled rather than stored?
- Why do backward-looking benchmarks underestimate LLM scientific value?
- What concrete problems do LLMs solve at the computational level?
- Does LLM miscalibration cause failures in clinical information extraction?
- How do codebook length and data similarity affect accuracy in LLM coding?
- How does intersubjective validation differ from pattern recognition in training data?
- How can we probe LLM representations in channels that training did not target?
- Does scaling model size solve compositional generalization problems?
- What power-law scaling patterns emerge when consistency models are trained at scale?
- How do parameter scaling and latent vectors interact in language models?
- Can recursion alone drive generalization better than model scale?
- Would hybrid systems combining LLMs with symbolic solvers overcome the retraction limitation?
- How do LLMs compress specific expert knowledge into median abstraction?
- How does algorithmic control flow define computational graph structure in LLM programs?
- Why does genetic programming outperform direct LLM generation by 86 percent?
- Can LLMs reliably generate novel working architectures without structured representations?
- Can optimization algorithms exploit the shift between procedural and planning bottlenecks?
- How many particles and iterations does optimal expert discovery require?
- What mechanism causes LLMs to plateau on numerical optimization tasks?
- Can LLMs successfully translate natural language into formal solver specifications?
- How should organizations redesign workflows if LLMs cannot solve optimization directly?
- Why do language models plateau at 55 to 60 percent constraint satisfaction?
- Why do LLMs fail at directly solving stochastic control problems?
- What latent mechanisms do LLMs use when they cannot execute iterative methods?
- What makes natural-language APIs particularly suited to LLM-based simulation?
- Can tool use or self-conditioning fix degradation in extended LLM workflows?
- Why do language models plateau at constraint satisfaction regardless of scale?
- Can smaller LLMs perform tool use tasks through modular decomposition?
- What prevents monolithic LLMs from coordinating decomposition with execution?
- Why do LLMs fail at iterative numerical computation in latent space?
- What constraint satisfaction rate do LLMs achieve at scale?
- Why do LLMs struggle more when only numerical values change?
- Can LLMs simultaneously reason and optimize their own modules?
- Why does fixing decomposition step count matter more than vocabulary alignment?
- How should LLM abstraction tools be evaluated without manual labeling?
- Does constraint-setting before generation change what LLMs can contribute?
- How does the discrete token bottleneck prevent gradient flow in language model control?
- Why do different LLMs converge on similar outputs in open-ended tasks?
- Why does most refinement in iterative models maintain answers rather than improve them?
- Why does iterative refinement amplify rather than correct reasoning errors?
- How does symbolic solver feedback differ from language-based self-critique?
- How does latent space diffusion enable evolutionary search in high dimensions?
- How does fitness-proportional selection guide LLM recombination in unstructured solution spaces?
- Can LLM-based crossover and mutation work in unstructured natural language spaces?
- How does the generation-verification gap shape evolutionary program search?
- Can evolutionary search at inference time scale beyond natural language planning?
- Can prompt optimization alone inject knowledge models don't already have?
- Can prompt optimization inject new knowledge into language models?
- Can prompt optimization inject genuinely new knowledge into a model?
- What knowledge can prompt optimization actually activate in trained models?
- Why does prompt optimization alone fail to inject genuinely new knowledge?
- Can latent recurrence and energy minimization both escape the same computational depth constraints?
- What distinguishes hierarchical dual-recurrence from flat parameter-sharing recurrence?
- Why does iterative refinement fail when information stays constant?
- Can width-scaling replace depth-scaling on inherently sequential problems?
- Can a two-layer network outgeneralize billion-parameter models through recursion alone?
- Can a tiny recursive network beat billion-parameter models on hard problems?
- Do iterative refinement methods reproduce the same overthinking failure mode?
- Can neural networks implement genuine algorithms or only statistical pattern matching?
- How do normalization and input injection control emergence of fixed points?
- Why do large language models outperform fine-tuned models once repeated items are removed?
- How do out-of-distribution tests reveal that optimization learning is memorization?
- Can critique-only calls in LLMs exploit a measurable gap between generation and evaluation?
- Can instruction prompts reliably steer an LLM judge toward specific alignment targets?
- Do small models show different parameter efficiency patterns than large models?
- How should tiny language models be architected differently than large ones?
- What filtering criteria best identify student-compatible refinements from teacher models?
- Can smaller language models match large model accuracy at lower total cost?
- Does sequence prediction accuracy prove an underlying world model exists?
- Can trained models encode programs more complex than their data-generating process?
- Why does extended chain-of-thought reasoning fail to improve numerical optimization performance?
- Why do reasoning models fail to improve constrained optimization performance?
- Can surface-level correctness hide failures in structural learning by LLMs?
- Can measuring answer-space collapse show LLM narrowing in practice?
- Do independent LLM outputs converge enough to create artificial hiveminds?
- How much do different LLMs independently converge on similar outputs?
- Do different large language models independently converge on identical outputs?
- Why do LLM groups match human aggregate patterns without similar processes?
Related concepts in this collection 5
This note in its neighbourhood — explore the map, then jump to a related concept in the list below.
Concept map
Do large language models actually perform iterat…
Do larger language models solve constrained optimi…
Do reasoning models actually beat standard models …
Do fine-tuned language models actually learn optim…
Does chain-of-thought reasoning reveal genuine inf…
What do models actually learn from chain-of-though…
Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph
your link
semantically near
linked from elsewhere
-
Do larger language models solve constrained optimization better?
Explores whether scaling LLMs—through more parameters, better training, or reasoning extensions—improves their ability to satisfy constraints in real optimization problems like power grids and portfolios.
same paper, the plateau this mechanism explains
-
Do reasoning models actually beat standard models on optimization?
Explores whether extended chain-of-thought in reasoning models delivers performance gains on constraint-satisfaction problems like power-grid optimization. Matters because reasoning models are treated as automatic upgrades, but the evidence may not support that claim.
same paper, why extended CoT does not fix it
-
Do fine-tuned language models actually learn optimization procedures?
Can RL fine-tuning teach LLMs to solve constraint-optimization problems through genuine reasoning, or does it merely sharpen pattern-matching? Testing on out-of-distribution variants reveals the mechanism.
same paper, the diagnostic that exposes the memorization
-
Does chain-of-thought reasoning reveal genuine inference or pattern matching?
Explores whether CoT instructions unlock real reasoning capabilities or simply constrain models to mimic familiar reasoning patterns from training data. This matters for understanding whether language models can actually reason abstractly.
adjacent: imitation vs computation
-
What do models actually learn from chain-of-thought training?
When models train on reasoning demonstrations, do they memorize content details or absorb reasoning structure? Testing with corrupted data reveals which aspects of CoT samples actually drive learning.
adjacent: form-over-content failure mode
Related papers in this collection 8
Papers most semantically related to this note, ranked by cosine similarity in the embedding space.
- Branch-Solve-Merge Improves Large Language Model Evaluation and Generation
- Can Large Language Models Reason and Optimize Under Constraints?
- Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language Models
- Metacognitive Reuse: Turning Recurring LLM Reasoning Into Concise Behaviors
- Chain of Thoughtlessness? An Analysis of CoT in Planning
- FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance
- Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models -- A Survey
- A Mechanistic Analysis of Looped Reasoning Language Models
Original note title
LLMs cannot execute iterative numerical methods in latent space and fall back to result guessing against memorized templates