The upper-tail accumulation scale derived from the gap-counting function N_n sets the critical inverse temperature for softmax attention concentration, unifying prior conflicting laws as special cases of different N_n.
hub
RoFormer: Enhanced Transformer with Rotary Position Embedding
68 Pith papers cite this work. Polarity classification is still indexing.
68
Pith papers citing it
abstract
Position encoding recently has shown effective in the transformer architecture. It enables valuable supervision for dependency modeling between elements at different positions of the sequence. In this paper, we first investigate various methods to integrate positional information into the learning process of transformer-based language models. Then, we propose a novel method named Rotary Position Embedding(RoPE) to effectively leverage the positional information. Specifically, the proposed RoPE encodes the absolute position with a rotation matrix and meanwhile incorporates the explicit relative position dependency in self-attention formulation. Notably, RoPE enables valuable properties, including the flexibility of sequence length, decaying inter-token dependency with increasing relative distances, and the capability of equipping the linear self-attention with relative position encoding. Finally, we evaluate the enhanced transformer with rotary position embedding, also called RoFormer, on various long text classification benchmark datasets. Our experiments show that it consistently overcomes its alternatives. Furthermore, we provide a theoretical analysis to explain some experimental results. RoFormer is already integrated into Huggingface: \url{https://huggingface.co/docs/transformers/model_doc/roformer}.
hub tools
citation-role summary
background 2
method 1
citation-polarity summary
claims ledger
- abstract Position encoding recently has shown effective in the transformer architecture. It enables valuable supervision for dependency modeling between elements at different positions of the sequence. In this paper, we first investigate various methods to integrate positional information into the learning process of transformer-based language models. Then, we propose a novel method named Rotary Position Embedding(RoPE) to effectively leverage the positional information. Specifically, the proposed RoPE encodes the absolute position with a rotation matrix and meanwhile incorporates the explicit relative
co-cited works
representative citing papers
RULER shows most long-context LMs drop sharply in performance on complex tasks as length and difficulty increase, with only half maintaining results at 32K tokens.
Mamba is a linear-time sequence model using input-dependent selective SSMs that achieves SOTA results across modalities and matches twice-larger Transformers on language modeling with 5x higher inference throughput.
Score entropy loss enables discrete diffusion models (SEDD) that cut perplexity 25-75% versus prior diffusion methods and outperform GPT-2 on language modeling while supporting infilling and compute-quality tradeoffs.
CRePE supplies depth-aware positional distributions along curved rays for stable unified-camera control in frozen video DiT models.
Embedding Temporal Logic enables runtime monitoring of temporally extended perceptual behaviors by defining predicates via distances between observed and reference embeddings in learned spaces, with conformal calibration for reliable evaluation.
Chirality emerges in SMILES translation models through an abrupt encoder-centered reorganization of representations after a long plateau, identified via checkpoint analysis and ablation.
CGAD is a staleness-aware Adam variant for DiLoCo that gates gradients with cosine and exponential decay, proves a convergence bound independent of maximum delay, and demonstrates stable pretraining of 25M to 7B parameter Llama-style models across controlled delays.
Jordan-RoPE realizes a non-semisimple relative positional operator that produces coupled oscillatory-polynomial features such as d e^{i omega d} for causal query-key lags.
Graph transformer RL for dynamic RMSA supports up to 13% more traffic than benchmarks on networks up to 143 nodes and 362 links.
A single end-to-end Transformer model unifies stellar labels from heterogeneous spectroscopic surveys into a self-consistent scale without post-hoc recalibration.
Physics-informed transformer with sin^2(theta) encoding, physics-aware positional encoding, multi-task decoder, and three-stage curriculum classifies powder diffraction into 99 extinction groups, with structured errors on symmetry subgroup hierarchy.
A Semantic Progress Function is defined as a 1D curve of cumulative semantic shifts from frame embeddings, supporting a linearization procedure that retimes video sequences for constant-rate semantic evolution.
WildSplatter jointly learns 3D Gaussians and appearance embeddings from unconstrained photo collections to enable fast feed-forward reconstruction and flexible lighting control in 3D Gaussian Splatting.
The work demonstrates masked-token prediction with transformers for model-independent anomaly detection in LHC data, achieving strong results on top-rich BSM signatures like four-top production using VQ-VAE tokenization.
FP32-converged language models enter a post-convergence phase where INT4 quantization error explodes while FP32 perplexity remains stable, with onset tied to fine convergence rather than learning rate decay.
Dual Triangle Attention achieves effective bidirectional attention with built-in positional inductive bias via dual triangular masks, outperforming standard bidirectional attention on position-sensitive tasks and showing strong masked language modeling results with or without positional embeddings.
A recurrent-depth architecture enables language models to improve reasoning performance by iterating computation in latent space, achieving gains equivalent to much larger models on benchmarks.
Transformers and SSMs are unified through structured state space duality, producing a 2-8X faster Mamba-2 model that remains competitive with Transformers.
A 2x growth factor in model warmstarting yields reliable training speedups for language models under 20 tokens/parameter budgets, with an empirical upper bound on effective growth factors.
PMNet uses unitary phasor dynamics and hierarchical anchors to make explicit memory stable for long sequences, matching a 3x larger Mamba model on long-context robustness with a 119M parameter network.
SpaceMind++ adds an explicit voxelized allocentric cognitive map and coordinate-guided fusion to video MLLMs, claiming SOTA on VSI-Bench and improved out-of-distribution generalization on three other 3D benchmarks.
Transformer components arise as the natural solution to precision-weighted directional state estimation on the hypersphere.
Looped MoE models scale better than standard transformers because different experts activate on each loop pass, recovering expressivity without extra parameters, and support superior early exits.
citing papers explorer
-
A Unified Framework for Critical Scaling of Inverse Temperature in Self-Attention
The upper-tail accumulation scale derived from the gap-counting function N_n sets the critical inverse temperature for softmax attention concentration, unifying prior conflicting laws as special cases of different N_n.
-
RULER: What's the Real Context Size of Your Long-Context Language Models?
RULER shows most long-context LMs drop sharply in performance on complex tasks as length and difficulty increase, with only half maintaining results at 32K tokens.
-
Mamba: Linear-Time Sequence Modeling with Selective State Spaces
Mamba is a linear-time sequence model using input-dependent selective SSMs that achieves SOTA results across modalities and matches twice-larger Transformers on language modeling with 5x higher inference throughput.
-
Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
Score entropy loss enables discrete diffusion models (SEDD) that cut perplexity 25-75% versus prior diffusion methods and outperform GPT-2 on language modeling while supporting infilling and compute-quality tradeoffs.
-
CRePE: Curved Ray Expectation Positional Encoding for Unified-Camera-Controlled Video Generation
CRePE supplies depth-aware positional distributions along curved rays for stable unified-camera control in frozen video DiT models.
-
Runtime Monitoring of Perception-Based Autonomous Systems via Embedding Temporal Logic
Embedding Temporal Logic enables runtime monitoring of temporally extended perceptual behaviors by defining predicates via distances between observed and reference embeddings in learned spaces, with conformal calibration for reliable evaluation.
-
From Syntax to Semantics: Unveiling the Emergence of Chirality in SMILES Translation Models
Chirality emerges in SMILES translation models through an abrupt encoder-centered reorganization of representations after a long plateau, identified via checkpoint analysis and ablation.
-
Cosine-Gated Adam-Decay: Drop-In Staleness-Aware Outer Optimization for Decoupled DiLoCo
CGAD is a staleness-aware Adam variant for DiLoCo that gates gradients with cosine and exponential decay, proves a convergence bound independent of maximum delay, and demonstrates stable pretraining of 25M to 7B parameter Llama-style models across controlled delays.
-
Jordan-RoPE: Non-Semisimple Relative Positional Encoding via Complex Jordan Blocks
Jordan-RoPE realizes a non-semisimple relative positional operator that produces coupled oscillatory-polynomial features such as d e^{i omega d} for causal query-key lags.
-
Graph Transformers and Stabilized Reinforcement Learning for Large-Scale Dynamic Routing Modulation and Spectrum Allocation in Elastic Optical Networks
Graph transformer RL for dynamic RMSA supports up to 13% more traffic than benchmarks on networks up to 143 nodes and 362 links.
-
Homogeneous Stellar Parameters from Heterogeneous Spectra with Deep Learning
A single end-to-end Transformer model unifies stellar labels from heterogeneous spectroscopic surveys into a self-consistent scale without post-hoc recalibration.
-
Attention Is Not All You Need for Diffraction
Physics-informed transformer with sin^2(theta) encoding, physics-aware positional encoding, multi-task decoder, and three-stage curriculum classifies powder diffraction into 99 extinction groups, with structured errors on symmetry subgroup hierarchy.
-
Video Analysis and Generation via a Semantic Progress Function
A Semantic Progress Function is defined as a 1D curve of cumulative semantic shifts from frame embeddings, supporting a linearization procedure that retimes video sequences for constant-rate semantic evolution.
-
WildSplatter: Feed-forward 3D Gaussian Splatting with Appearance Control from Unconstrained Images
WildSplatter jointly learns 3D Gaussians and appearance embeddings from unconstrained photo collections to enable fast feed-forward reconstruction and flexible lighting control in 3D Gaussian Splatting.
-
Masked-Token Prediction for Anomaly Detection at the Large Hadron Collider
The work demonstrates masked-token prediction with transformers for model-independent anomaly detection in LHC data, achieving strong results on top-rich BSM signatures like four-top production using VQ-VAE tokenization.
-
When Flat Minima Fail: Characterizing INT4 Quantization Collapse After FP32 Convergence
FP32-converged language models enter a post-convergence phase where INT4 quantization error explodes while FP32 perplexity remains stable, with onset tied to fine convergence rather than learning rate decay.
-
Dual Triangle Attention: Effective Bidirectional Attention Without Positional Embeddings
Dual Triangle Attention achieves effective bidirectional attention with built-in positional inductive bias via dual triangular masks, outperforming standard bidirectional attention on position-sensitive tasks and showing strong masked language modeling results with or without positional embeddings.
-
Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
A recurrent-depth architecture enables language models to improve reasoning performance by iterating computation in latent space, achieving gains equivalent to much larger models on benchmarks.
-
Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality
Transformers and SSMs are unified through structured state space duality, producing a 2-8X faster Mamba-2 model that remains competitive with Transformers.
-
When is Warmstarting Effective for Scaling Language Models?
A 2x growth factor in model warmstarting yields reliable training speedups for language models under 20 tokens/parameter budgets, with an empirical upper bound on effective growth factors.
-
Phasor Memory Networks: Stable Backpropagation Through Time for Scalable Explicit Memory
PMNet uses unitary phasor dynamics and hierarchical anchors to make explicit memory stable for long sequences, matching a 3x larger Mamba model on long-context robustness with a 119M parameter network.
-
SpaceMind++: Toward Allocentric Cognitive Maps for Spatially Grounded Video MLLMs
SpaceMind++ adds an explicit voxelized allocentric cognitive map and coordinate-guided fusion to video MLLMs, claiming SOTA on VSI-Bench and improved out-of-distribution generalization on three other 3D benchmarks.
-
RT-Transformer: The Transformer Block as a Spherical State Estimator
Transformer components arise as the natural solution to precision-weighted directional state estimation on the hypersphere.
-
Sparse Layers are Critical to Scaling Looped Language Models
Looped MoE models scale better than standard transformers because different experts activate on each loop pass, recovering expressivity without extra parameters, and support superior early exits.
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception is a training-free method that injects latent KV banks at chosen layers to steer LLMs, achieving superior control-drift balance and up to 118x storage reduction on personality and structured-reasoning tasks.
-
Spectral Lens: Activation and Gradient Spectra as Diagnostics of LLM Optimization
Spectral analysis of activations and gradients provides new diagnostics that link batch size to representation geometry, early covariance tails to token efficiency, and spectral shifts to learning dynamics in decoder-only LLMs, backed by a mechanistic model.
-
Feature Starvation as Geometric Instability in Sparse Autoencoders
Adaptive elastic net SAEs (AEN-SAEs) mitigate feature starvation in SAEs by combining ℓ2 structural stability with adaptive ℓ1 reweighting, producing a Lipschitz-continuous sparse coding map that recovers global feature support under mild assumptions.
-
Spiking Sequence Machines and Transformers
Spiking SDM and transformers implement identical functional operations for sequences via cosine similarity retrieval, unified by a phase-latency isomorphism between spike timing and sinusoidal positional encoding.
-
Mutual Forcing: Dual-Mode Self-Evolution for Fast Autoregressive Audio-Video Character Generation
Mutual Forcing trains a single native autoregressive audio-video model with mutually reinforcing few-step and multi-step modes via self-distillation to match 50-step baselines at 4-8 steps.
-
Foveated Reasoning: Stateful, Action-based Visual Focusing for Vision-Language Models
Foveated Reasoner integrates foveation as stateful actions inside the autoregressive decoding loop of vision-language models, trained via cold-start supervision then reinforcement learning to achieve higher accuracy at low token budgets.
-
HypEHR: Hyperbolic Modeling of Electronic Health Records for Efficient Question Answering
HypEHR is a hyperbolic embedding model for EHR data that uses Lorentzian geometry and hierarchy-aware pretraining to answer clinical questions nearly as well as large language models but with much smaller size.
-
LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model
LLaDA2.0-Uni unifies multimodal understanding and generation inside one discrete diffusion large language model with a semantic tokenizer, MoE backbone, and diffusion decoder.
-
Towards Real-Time ECG and EMG Modeling on $\mu$NPUs
PhysioLite delivers Transformer-comparable ECG/EMG performance using learnable wavelet filters and hardware-aware design at ~370KB quantized size on μNPUs.
-
Shape: A Self-Supervised 3D Geometry Foundation Model for Industrial CAD Analysis
A 10.9M-parameter self-supervised model pretrained on 61k CAD meshes achieves R²=0.729 reconstruction and 98.1% top-1 retrieval on held-out data via masked normalized geometry reconstruction and multi-resolution contrastive learning.
-
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.
-
MT-OSC: Path for LLMs that Get Lost in Multi-Turn Conversation
MT-OSC condenses chat history via a one-off sequential process with a few-shot Condenser and lightweight Decider to reduce tokens and preserve LLM accuracy in multi-turn settings.
-
LoMa: Local Feature Matching Revisited
Scaling data, model size, and compute for local feature matching produces large performance gains on challenging benchmarks and a new manually annotated HardMatch dataset.
-
GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
GLM-4.5V reaches state-of-the-art results on 42 multimodal benchmarks among open-source models of similar size by applying reinforcement learning with curriculum sampling to a strong vision foundation model.
-
SAM 2: Segment Anything in Images and Videos
SAM 2 delivers more accurate video segmentation with 3x fewer user interactions and 6x faster image segmentation than the original SAM by training a streaming-memory transformer on the largest video segmentation dataset collected to date.
-
Chameleon: Mixed-Modal Early-Fusion Foundation Models
Chameleon is an early-fusion token model that handles mixed image-text sequences for understanding and generation, achieving competitive or superior performance to larger models like Llama-2, Mixtral, and Gemini-Pro on captioning, VQA, text, and image tasks.
-
StarCoder 2 and The Stack v2: The Next Generation
StarCoder2-15B matches or beats CodeLlama-34B on code tasks despite being smaller, and StarCoder2-3B outperforms prior 15B models, with open weights and exact training data identifiers released.
-
Efficient Streaming Language Models with Attention Sinks
StreamingLLM lets finite-window LLMs generalize to infinite-length sequences by retaining initial-token KV states as attention sinks, enabling stable streaming inference up to 4M tokens.
-
YaRN: Efficient Context Window Extension of Large Language Models
YaRN extends the context window of RoPE-based LLMs like LLaMA more efficiently than prior methods, using 10x fewer tokens and 2.5x fewer steps while surpassing state-of-the-art performance and enabling extrapolation beyond fine-tuning lengths.
-
Retentive Network: A Successor to Transformer for Large Language Models
RetNet is a new sequence modeling architecture that delivers parallel training, constant-time inference, and competitive language modeling performance as a potential replacement for Transformers.
-
Textbooks Are All You Need
A 1.3B-parameter code model trained on 7B tokens of curated textbook and synthetic data achieves 50.6% on HumanEval, indicating data quality can enable strong performance at small scale.
-
BloombergGPT: A Large Language Model for Finance
BloombergGPT is a 50B parameter LLM trained on a 708B token mixed financial and general dataset that outperforms prior models on financial benchmarks while preserving general LLM performance.
-
PaLM: Scaling Language Modeling with Pathways
PaLM 540B demonstrates continued scaling benefits by setting new few-shot SOTA results on hundreds of benchmarks and outperforming humans on BIG-bench.
-
Mela: Test-Time Memory Consolidation based on Transformation Hypothesis
Mela is a Transformer variant with a dual-frequency Hierarchical Memory Module and MemStack that performs test-time memory consolidation, outperforming baselines on long contexts.
-
StreamIndex: Memory-Bounded Compressed Sparse Attention via Streaming Top-k
Chunked streaming top-k enables CSA indexer execution at 1M sequence length with 6.21 GB peak memory and >=0.998 recall on synthetic V4-shaped inputs.
-
Nautile-370M: Spectral Memory Meets Attention in a Small Reasoning Model
Nautile-370M is a hybrid small language model using SeqCond Attention layers alternating with transformers, with a claimed proof that the spectral operator matches full self-attention expressiveness in the continuous limit.