Chain-of-thought prompting, by including intermediate reasoning steps in few-shot examples, elicits strong reasoning abilities in large language models on arithmetic, commonsense, and symbolic tasks.
super hub Canonical reference
Scaling Laws for Neural Language Models
Canonical reference. 84% of citing Pith papers cite this work as background.
845
Pith papers citing it
Background
84% of classified citations
abstract
We study empirical scaling laws for language model performance on the cross-entropy loss. The loss scales as a power-law with model size, dataset size, and the amount of compute used for training, with some trends spanning more than seven orders of magnitude. Other architectural details such as network width or depth have minimal effects within a wide range. Simple equations govern the dependence of overfitting on model/dataset size and the dependence of training speed on model size. These relationships allow us to determine the optimal allocation of a fixed compute budget. Larger models are significantly more sample-efficient, such that optimally compute-efficient training involves training very large models on a relatively modest amount of data and stopping significantly before convergence.
hub tools
citation-role summary
background 121
method 6
dataset 3
baseline 2
other 2
citation-polarity summary
claims ledger
- abstract We study empirical scaling laws for language model performance on the cross-entropy loss. The loss scales as a power-law with model size, dataset size, and the amount of compute used for training, with some trends spanning more than seven orders of magnitude. Other architectural details such as network width or depth have minimal effects within a wide range. Simple equations govern the dependence of overfitting on model/dataset size and the dependence of training speed on model size. These relationships allow us to determine the optimal allocation of a fixed compute budget. Larger models are s
authors
co-cited works
representative citing papers
BBO-Pile is the first large-scale open dataset of real optimization trajectories used to train and scale foundation models that imitate black-box optimization methods.
Introduces the Synthetic Data Contamination Equilibrium and derives closed-form optimal provenance subsidies s* = KL(q||p)/(2 kappa) plus watermark strengths to mitigate model collapse, validated by OLS matching structural predictions on C4 data.
Cayley unitary adapters executed on real quantum hardware improve LLM perplexity by 1.4% on Llama 3.1 8B with 6000 parameters and recover 83% of compression-induced degradation on SmolLM2.
ε-coresets for attention exist of size O(√d e^{ρ+o(ρ)}/ε) for unit-norm keys/values and queries of norm ≤ρ, nearly matching the Ω(√d e^ρ/ε) lower bound.
RoundPipe achieves near-zero-bubble pipeline parallelism for LLM training on consumer GPUs by dynamically dispatching computation stages round-robin, yielding 1.48-2.16x speedups and enabling 235B model fine-tuning on 8x RTX 4090.
Transformer weight spectra exhibit transient compression waves that propagate layer-wise, persistent non-monotonic depth gradients in power-law exponents, and Q/K-V asymmetry, with the spectral exponent alpha predicting layer importance and enabling pruning gains of 1.1x-3.6x over Last-N baselines.
Grokking reflects escape from a metastable low-dimensional regime where transverse curvature accumulates before generalization, with subspace motion necessary but curvature boost insufficient.
The SDE benchmark shows LLMs lag on scientific discovery tasks relative to general science tests, with diminishing scaling returns and shared weaknesses across models.
Transformers perform kernel-based prediction for Hölder regression on manifolds and achieve intrinsic-dimension-dependent minimax rates with sufficient training tasks.
Introduces hybrid noise and novel coupling analysis to achieve the first convergent hidden-state DP bound for zeroth-order optimization.
LLaDA is a scalable diffusion-based language model that matches autoregressive LLMs like LLaMA3 8B on tasks and surpasses GPT-4o on reversal poem completion.
ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.
TTT layers treat the hidden state as a trainable model updated at test time, allowing linear-complexity sequence models to scale perplexity reduction with context length unlike Mamba.
KANs with learnable univariate spline activations on edges achieve better accuracy than MLPs with fewer parameters, faster scaling, and direct visualization for scientific discovery.
Tiny language models under 10M parameters trained on a synthetic children's story dataset generate fluent, consistent, multi-paragraph English text with near-perfect grammar and reasoning.
Pythia releases 16 identically trained LLMs with full checkpoints and data tools to study training dynamics, scaling, memorization, and bias in language models.
Language models can automatically generate high-quality evaluation datasets that reveal new cases of inverse scaling, sycophancy, and concerning goal-seeking behaviors, including some worsened by RLHF.
The Pile is a newly constructed 825 GiB dataset from 22 diverse sources that enables language models to achieve better performance on academic, professional, and cross-domain tasks than models trained on Common Crawl variants.
HERMES provides a reusable hierarchical labeling substrate for pre-training data that reveals granularity-specific effects in data mixing rules during model training.
Multi-agent LLMs generate and verify 14,073 deterministic reaction rules from 665,901 patents, enabling 97.7% classification of unseen reactions with finer resolution than fixed proprietary systems.
SemRF supplies fixed semantic anchors and pseudo-inverse tying to produce stable coordinates for residual dynamics, Voronoi traces, and minimum-action canonical paths that link to parameter efficiency under controlled interface error.
Cortex uses an Ontological Corpus Graph to structure web-scale corpora, creating a refined 24.14B-token corpus and a new benchmark validated on eight LLMs.
Token loss trajectories follow localized sigmoids whose learning-time spectrum quantitatively reconstructs scaling-law derivatives on T, D, and M axes and enables faster training via distribution reshaping.
citing papers explorer
-
Prescriptive Scaling Laws for Data Constrained Training
A one-parameter scaling law models excess loss from data repetition as an additive overfitting penalty, recommending model capacity increases over excessive repetition and showing that strong weight decay reduces the penalty coefficient by ~70%.
-
Importance-Guided Basis Selection for Low-Rank Decomposition of Large Language Models
BSI ranks singular-vector bases for LLM low-rank compression by estimating expected task loss increase via second-order Taylor expansion of the loss and an efficient Hessian-diagonal estimator, outperforming magnitude-based baselines on math reasoning benchmarks.
-
Compute Optimal Tokenization
In compute-optimal regimes, language model parameter count scales proportionally with data bytes rather than tokens, and the optimal compression rate decreases with increasing compute.
-
When Less is Enough: Efficient Inference via Collaborative Reasoning
A large model generates a compact reasoning signal that a small model uses to solve tasks, reducing the large model's output tokens by up to 60% on benchmarks like AIME and GPQA.
-
LLM Ghostbusters: Surgical Hallucination Suppression via Adaptive Unlearning
Adaptive Unlearning suppresses package hallucinations in code-generating LLMs by 81% while preserving benchmark performance, using model-generated data and no human labels.
-
CLEAR: Revealing How Noise and Ambiguity Degrade Reliability in LLMs for Medicine
CLEAR reveals that LLMs' accuracy on medical questions drops and their 'humility deficit' grows as the number of plausible answers increases and abstention options shift from assertive to uncertain phrasing.
-
Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs
PVM adds a parallel branch to LVLMs that directly supplies visual embeddings to prevent attention decay over long generated sequences, yielding accuracy gains on reasoning tasks with minimal overhead.
-
Possibilistic Predictive Uncertainty for Deep Learning
DAPPr projects a possibilistic posterior over network parameters to predictions using supremum operators and approximates it with learnable Dirichlet functions to yield an efficient training objective for epistemic uncertainty.
-
AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs
AGoQ delivers up to 52% lower memory use and 1.34x faster training for 8B-32B LLaMA models by using near-4-bit adaptive activations and 8-bit gradients while preserving pretraining convergence and downstream accuracy.
-
Comprehensive AI governance requires addressing non-model gains
Non-model gains via inference, systems, and assets can drive AI capabilities independently of base models, requiring governance beyond model-level evaluation and mitigation.
-
InvEvolve: Evolving White-Box Inventory Policies via Large Language Models with Performance Guarantees
InvEvolve evolves inventory policies using LLMs with RL and provides statistical safety guarantees, outperforming classical and DL methods on synthetic and real data.
-
State Stream Transformer (SST) V2: Parallel Training of Nonlinear Recurrence for Latent Space Reasoning
SST V2 introduces parallel-trainable nonlinear recurrence in latent space to let transformers reason continuously across positions, delivering +15 points on GPQA-Diamond and halving remaining GSM8K errors over matched baselines.
-
Dynamic Cluster Data Sampling for Efficient and Long-Tail-Aware Vision-Language Pre-training
DynamiCS dynamically scales semantic clusters per training epoch to reduce VLM pre-training compute while improving accuracy on long-tail concepts compared to static or flattening baselines.
-
COPUS: Co-adaptive Parallelism and Batch Size Selection in Large Language Model Training
COPUS co-adapts batch size and parallelism during LLM training via goodput to deliver 3.9-8% average faster convergence than fixing one while tuning the other.
-
Mixture of Heterogeneous Grouped Experts for Language Modeling
MoHGE achieves standard MoE performance with 20% fewer parameters and balanced GPU utilization via grouped heterogeneous experts, two-level routing, and specialized auxiliary losses.
-
The Power of Power Law: Asymmetry Enables Compositional Reasoning
Power-law data sampling creates beneficial asymmetry in the loss landscape that lets models acquire high-frequency skill compositions first, enabling more efficient learning of rare long-tail skills than uniform distributions.
-
Adaptive Test-Time Compute Allocation with Evolving In-Context Demonstrations
An adaptive test-time framework uses a warm-up phase on the test set to build evolving in-context examples, then concentrates compute on unresolved queries to outperform static baselines on math, coding, and reasoning tasks with lower total inference cost.
-
Hybrid Policy Distillation for LLMs
Hybrid Policy Distillation unifies existing knowledge distillation methods for LLMs into a reweighted log-likelihood objective and introduces a hybrid forward-reverse KL approach with mixed data sampling to improve stability, efficiency, and performance.
-
Normalizing Flows with Iterative Denoising
iTARFlow augments normalizing flows with diffusion-style iterative denoising during sampling while preserving end-to-end likelihood training, reaching competitive results on ImageNet 64/128/256.
-
Volume Transformer: Revisiting Vanilla Transformers for 3D Scene Understanding
A minimally modified vanilla Transformer called Volt achieves state-of-the-art 3D semantic and instance segmentation by using volumetric tokens, 3D rotary embeddings, and a data-efficient training recipe that scales better than domain-specific backbones.
-
Understanding the Mechanism of Altruism in Large Language Models
A small set of sparse autoencoder features in LLMs drives shifts between generous and selfish allocations in dictator games, with causal patching and steering confirming their role and generalization to other social games.
-
SAMoRA: Semantic-Aware Mixture of LoRA Experts for Task-Adaptive Learning
SAMoRA is a parameter-efficient fine-tuning framework that uses semantic-aware routing and task-adaptive scaling within a Mixture of LoRA Experts to improve multi-task performance and generalization over prior methods.
-
OmniMouse: Scaling properties of multi-modal, multi-task Brain Models on 150B Neural Tokens
OmniMouse demonstrates data-driven scaling in multi-task brain models on a 150B-token neural dataset, achieving SOTA across prediction, decoding, and forecasting while model size gains saturate.
-
Measuring Distribution Shift in User Prompts and Its Effects on LLM Performance
The LENS framework applied to 192 real-world settings shows moderate natural prompt distribution shifts cause 73% average performance loss in deployed LLMs, especially across user groups and regions.
-
Hive: A Multi-Agent Infrastructure for Algorithm- and Task-Level Scaling
Hive is a multi-agent infrastructure with a logits cache for reducing cross-path redundancy in sampling and agent-aware scheduling for better compute and KV-cache allocation, shown to deliver 1.11x-1.76x speedups and 33%-51% lower hotspot miss rates.
-
Probabilistic Programs of Thought
Probabilistic programs of thought let LLMs produce many program variants from one generation by building a compact probabilistic representation of the token distribution.
-
MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search
MemSearch-o1 mitigates memory dilution in agentic LLM search through reasoning-aligned token-level memory growth, retracing with a contribution function, and path reorganization, improving reasoning activation on benchmarks.
-
Graph-Guided Adaptive Channel Elimination for KV Cache Compression
GRACE reframes KV cache channel pruning as graph optimization to find a near-optimal subset, achieving 60% compression with negligible degradation and outperforming prior methods.
-
Application of a Mixture of Experts-based Foundation Model to the GlueX DIRC Detector
A single MoE-based foundation model with transformer backbone unifies simulation, PID, and noise filtering for the GlueX DIRC detector and matches or exceeds traditional geometrical and prior deep-learning methods across kinematics.
-
TRON: Trainable, architecture-reconfigurable random optical neural networks
TRON demonstrates a trainable and reconfigurable optical neural network that combines multi-scattering media with DMD-based matrix multiplication and performs in-situ optimization plus neural architecture search on the optical hardware itself.
-
Predicting Power-System Dynamic Trajectories with Foundation Models
LASS-ODE-Power is a pretrained model that predicts power-system dynamic trajectories across regimes in a zero-shot manner after large-scale ODE pretraining and targeted fine-tuning.
-
Adaptive Test-Time Compute Allocation for Reasoning LLMs via Constrained Policy Optimization
A Lagrangian-relaxation plus imitation-learning pipeline adaptively allocates test-time compute to LLMs, outperforming uniform baselines by up to 12.8% relative accuracy on MATH while staying within a fixed average budget.
-
ELMoE-3D: Leveraging Intrinsic Elasticity of MoE for Hybrid-Bonding-Enabled Self-Speculative Decoding in On-Premises Serving
ELMoE-3D achieves 6.6x average speedup and 4.4x energy efficiency gain for MoE serving on 3D hardware by scaling expert and bit elasticity for elastic self-speculative decoding.
-
Mistake gating leads to energy and memory efficient continual learning
Mistake-gated plasticity reduces neural network updates by 50-80% by gating changes on classification errors, improving efficiency for continual learning without added hyperparameters.
-
Evaluation of Agents under Simulated AI Marketplace Dynamics
Marketplace Evaluation uses repeated-interaction simulations to assess information access systems with marketplace-level metrics such as retention and market share that complement traditional accuracy measures.
-
XRZero-G0: Pushing the Frontier of Dexterous Robotic Manipulation with Interfaces, Quality and Ratios
XRZero-G0 enables 2000-hour robot-free datasets that, when mixed 10:1 with real-robot data, match full real-robot performance at 1/20th the cost and support zero-shot transfer.
-
Efficient Retrieval Scaling with Hierarchical Indexing for Large Scale Recommendation
A jointly learned hierarchical index with cross-attention and residual quantization scales exact retrieval in foundational recommendation models, deployed at Meta with additional performance from test-time training on index nodes.
-
Parcae: Scaling Laws For Stable Looped Language Models
Parcae stabilizes looped LLMs via spectral norm constraints on injection parameters, enabling power-law scaling for training FLOPs and saturating exponential scaling at test time that improves quality over fixed-depth baselines under fixed parameter budgets.
-
Round-Trip Translation Reveals What Frontier Multilingual Benchmarks Miss
Round-trip translation evaluation shows that existing multilingual benchmarks measure reasoning and recall instead of language skills, with the new LiT benchmark correlating at rho=0.94 to LMArena ratings.
-
BID-LoRA: A Parameter-Efficient Framework for Continual Learning and Unlearning
BID-LoRA uses bi-directional low-rank adapters with retain/new/unlearn pathways and escape unlearning to enable continual learning and unlearning while minimizing knowledge leakage and parameter updates.
-
Adaptive Test-Time Scaling for Zero-Shot Respiratory Audio Classification
TRIAGE adaptively scales test-time compute via tiered zero-shot stages for respiratory audio classification, reaching mean AUROC 0.744 across nine tasks while outperforming prior zero-shot methods.
-
ReflectCAP: Detailed Image Captioning with Reflective Memory
ReflectCAP distills model-specific hallucination and oversight patterns into Structured Reflection Notes that steer LVLMs toward more factual and complete image captions, reaching the Pareto frontier on factuality-coverage trade-offs.
-
How Transformers Learn to Plan via Multi-Token Prediction
Multi-token prediction induces a two-stage reverse reasoning process in Transformers via gradient decoupling, improving planning on synthetic and realistic tasks.
-
Omnimodal Dataset Distillation via High-order Proxy Alignment
HoPA captures high-order cross-modal alignments via a shared proxy to enable scalable omnimodal dataset distillation with better performance-compression trade-offs.
-
Universal statistical signatures of evolution in artificial intelligence architectures
AI architectural modifications exhibit a heavy-tailed Student's t-distribution of fitness effects with 68% deleterious, 19% neutral, and 13% beneficial changes, closely matching distributions in D. melanogaster and S. cerevisiae.
-
Visual Enhanced Depth Scaling for Multimodal Latent Reasoning
Visual replay module and adaptive depth scaling improve multimodal latent reasoning, reaching SOTA benchmarks with faster inference than explicit chain-of-thought methods.
-
Pioneer Agent: Continual Improvement of Small Language Models in Production
Pioneer Agent automates the full lifecycle of adapting and continually improving small language models via diagnosis-driven data synthesis and regression-constrained retraining, delivering gains of 1.6-83.8 points on benchmarks and large lifts in production-style tasks.
-
Integrated electro-optic attention nonlinearities for transformers
Thin-film lithium niobate modulators implement electro-optic Softmax and Sigmoid alternatives for transformers that maintain competitive accuracy under 4-bit quantization and characterized noise up to 10 GBaud.
-
Visually-grounded Humanoid Agents
A coupled world-agent framework uses 3D Gaussian reconstruction and first-person RGB-D perception with iterative planning to enable goal-directed, collision-avoiding humanoid behavior in novel reconstructed scenes.
-
Scaling-Aware Data Selection for End-to-End Autonomous Driving Systems
MOSAIC is a scaling-aware data selection framework that outperforms baselines in training end-to-end autonomous driving planners, achieving comparable or better EPDMS scores with up to 80% less data.