Skip to content
← Back to explorer

Pretraining with Token-Level Adaptive Latent Chain-of-Thought

Boyi Zeng, Yiqin Hao, He Li, Shixiang Song, Feichen Song, Zitong Wang, Siyuan Huang, Yi Xu, ZiWei He, Xinbing Wang, Zhouhan Lin · Feb 9, 2026 · Citations: 0

Automatic Metrics General Long Horizon
Open arXiv Find Implementation RSS feed Shortlist (0)

Data freshness

Extraction: Fresh

Check recency before relying on this page for active eval decisions. Use stale pages as context and verify against current hub results.

Metadata refreshed

Mar 10, 2026, 5:18 AM

Recent

Extraction refreshed

Mar 13, 2026, 11:05 PM

Fresh

Extraction source

Persisted extraction

Confidence 0.45

Abstract

Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs. This work explores an alternative axis: increasing per-token computation without expanding parameters, by internalizing latent Chain-of-Thought (CoT) into pretraining. We propose Pretraining with Token-Level Adaptive Latent CoT (adaptive latent CoT), where the model generates a variable-length latent CoT trajectory before emitting each token -- allocating longer trajectories to difficult tokens and shorter (or even zero) trajectories to easy ones. Importantly, this behavior emerges naturally from one-stage pretraining on general text and reduces computation in both training and inference via token-wise adaptive halting. Experiments with Llama architectures show that adaptive latent CoT consistently improves language modeling perplexity and broad downstream accuracy, even with fewer training FLOPs than prior recurrent baselines.

Low-signal caution for protocol decisions

Use this page for context, then validate protocol choices against stronger HFEPX references before implementation decisions.

  • Extraction confidence is 0.45 (below strong-reference threshold).
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

HFEPX Relevance Assessment

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Best use

Background context only

Use if you need

A secondary eval reference to pair with stronger protocol papers.

Main weakness

Extraction confidence is 0.45 (below strong-reference threshold).

Trust level

Low

Eval-Fit Score

25/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Detected

HFEPX Fit

Adjacent candidate

Extraction confidence: Low

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Field Provenance & Confidence

Each key protocol field shows extraction state, confidence band, and data source so you can decide whether to trust it directly or validate from full text.

Human Feedback Types

missing

None explicit

Confidence: Low Source: Persisted extraction missing

No explicit feedback protocol extracted.

Evidence snippet: Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs.

Evaluation Modes

partial

Automatic Metrics

Confidence: Low Source: Persisted extraction evidenced

Includes extracted eval setup.

Evidence snippet: Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs.

Quality Controls

missing

Not reported

Confidence: Low Source: Persisted extraction missing

No explicit QC controls found.

Evidence snippet: Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs.

Benchmarks / Datasets

missing

Not extracted

Confidence: Low Source: Persisted extraction missing

No benchmark anchors detected.

Evidence snippet: Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs.

Reported Metrics

partial

Accuracy, Perplexity

Confidence: Low Source: Persisted extraction evidenced

Useful for evaluation criteria comparison.

Evidence snippet: Experiments with Llama architectures show that adaptive latent CoT consistently improves language modeling perplexity and broad downstream accuracy, even with fewer training FLOPs than prior recurrent baselines.

Rater Population

missing

Unknown

Confidence: Low Source: Persisted extraction missing

Rater source not explicitly reported.

Evidence snippet: Scaling large language models by increasing parameters and training data is increasingly constrained by limited high-quality corpora and rising communication costs.

Human Data Lens

  • Uses human feedback: No
  • Feedback types: None
  • Rater population: Unknown
  • Unit of annotation: Trajectory
  • Expertise required: General
  • Extraction source: Persisted extraction

Evaluation Lens

  • Evaluation modes: Automatic Metrics
  • Agentic eval: Long Horizon
  • Quality controls: Not reported
  • Confidence: 0.45
  • Flags: ambiguous

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

accuracyperplexity

Research Brief

Deterministic synthesis

We propose Pretraining with Token-Level Adaptive Latent CoT (adaptive latent CoT), where the model generates a variable-length latent CoT trajectory before emitting each token -- allocating longer trajectories to difficult tokens and… HFEPX signals include Automatic Metrics, Long Horizon with confidence 0.45. Updated from current HFEPX corpus.

Generated Mar 13, 2026, 11:05 PM · Grounded in abstract + metadata only

Key Takeaways

  • We propose Pretraining with Token-Level Adaptive Latent CoT (adaptive latent CoT), where the model generates a variable-length latent CoT trajectory before emitting each token --…
  • Experiments with Llama architectures show that adaptive latent CoT consistently improves language modeling perplexity and broad downstream accuracy, even with fewer training FLOPs…
  • Primary extracted protocol signals: Automatic Metrics, Long Horizon.

Researcher Actions

  • Treat this as method context, then pivot to protocol-specific HFEPX hubs.
  • Identify benchmark choices from full text before operationalizing conclusions.
  • Validate metric comparability (accuracy, perplexity).

Caveats

  • Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
  • Extraction confidence is probabilistic and should be validated for critical decisions.

Research Summary

Contribution Summary

  • We propose Pretraining with Token-Level Adaptive Latent CoT (adaptive latent CoT), where the model generates a variable-length latent CoT trajectory before emitting each token -- allocating longer trajectories to difficult tokens and…
  • Experiments with Llama architectures show that adaptive latent CoT consistently improves language modeling perplexity and broad downstream accuracy, even with fewer training FLOPs than prior recurrent baselines.

Researcher Checklist

  • Gap: Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Pass: Evaluation mode is explicit

    Detected: Automatic Metrics

  • Gap: Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Gap: Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Pass: Metric reporting is present

    Detected: accuracy, perplexity

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote