Difference between revisions of "AI research trends"
KevinYager (talk | contribs) (→Neural (non-token) Latent Representation) |
KevinYager (talk | contribs) (→Context Length) |
||
| (14 intermediate revisions by the same user not shown) | |||
| Line 30: | Line 30: | ||
* 2024-04-12: Meta et al. demonstrate [https://arxiv.org/abs/2404.08801 Megalodon] that enables infinite context via a more efficient architecture | * 2024-04-12: Meta et al. demonstrate [https://arxiv.org/abs/2404.08801 Megalodon] that enables infinite context via a more efficient architecture | ||
* 2024-04-14: Google presents [https://arxiv.org/abs/2404.09173 TransformerFAM], which leverages a feedback loop so it attends to its own latent representations, acting as working memory and provides effectively infinite context | * 2024-04-14: Google presents [https://arxiv.org/abs/2404.09173 TransformerFAM], which leverages a feedback loop so it attends to its own latent representations, acting as working memory and provides effectively infinite context | ||
| − | * [https://x.com/MiniMax__AI/status/1879226391352549451 2025- | + | * 2024-10-31: [https://arxiv.org/abs/2410.23771 What is Wrong with Perplexity for Long-context Language Modeling?] |
| − | * [https://x.com/Alibaba_Qwen/status/1883557964759654608 2025- | + | * [https://x.com/MiniMax__AI/status/1879226391352549451 2025-01-14]: [https://www.minimaxi.com/en/news/minimax-01-series-2 MiniMax-01] 4M ([https://www.minimaxi.com/en/news/minimax-01-series-2 paper]) |
| + | * [https://x.com/Alibaba_Qwen/status/1883557964759654608 2025-01-27]: [https://qwenlm.github.io/blog/qwen2.5-1m/ Qwen2.5-1M] ([https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen2.5-1M/Qwen2_5_1M_Technical_Report.pdf report]) | ||
| + | * 2025-02-14: [https://arxiv.org/abs/2502.08910 InfiniteHiP: Extending Language Model Context Up to 3 Million Tokens on a Single GPU] 3M | ||
| + | * [https://x.com/AnimaAnandkumar/status/1897449851941744648 2025-02-18]: [https://arxiv.org/abs/2502.12574 HeadInfer: Memory-Efficient LLM Inference by Head-wise Offloading] ([https://github.com/wdlctc/headinfer code]) | ||
| + | * 2025-02-18: [https://arxiv.org/abs/2502.12962 Infinite Retrieval: Attention Enhanced LLMs in Long-Context Processing] | ||
| + | * 2025-02-19: [https://github.com/MoonshotAI/MoBA MoBA: Mixture of Block Attention for Long-Context LLMs] | ||
| + | * 2025-02-27: [https://arxiv.org/abs/2502.20082 LongRoPE2: Near-Lossless LLM Context Window Scaling] ([https://github.com/microsoft/LongRoPE code]) | ||
==Extended Context== | ==Extended Context== | ||
| Line 49: | Line 55: | ||
=Updating Weights at Inference-time= | =Updating Weights at Inference-time= | ||
* 2025-01: [https://arxiv.org/abs/2501.06252 Transformer<sup>2</sup>: Self-adaptive LLMs] | * 2025-01: [https://arxiv.org/abs/2501.06252 Transformer<sup>2</sup>: Self-adaptive LLMs] | ||
| + | |||
| + | ==Parameters as Tokens== | ||
| + | * 2024-10: [https://arxiv.org/abs/2410.23168 TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parameters] ([https://github.com/Haiyang-W/TokenFormer code]) | ||
=Internal Thought Representation Space= | =Internal Thought Representation Space= | ||
| Line 65: | Line 74: | ||
* 2025-01: [https://arxiv.org/abs/2501.19201 Efficient Reasoning with Hidden Thinking] ([https://github.com/shawnricecake/Heima code]) | * 2025-01: [https://arxiv.org/abs/2501.19201 Efficient Reasoning with Hidden Thinking] ([https://github.com/shawnricecake/Heima code]) | ||
* 2025-02: [https://huggingface.co/tomg-group-umd/huginn-0125 Huginn-0125]: [https://arxiv.org/abs/2502.05171 Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach] ([https://github.com/seal-rg/recurrent-pretraining code], [https://huggingface.co/tomg-group-umd/huginn-0125 model]) | * 2025-02: [https://huggingface.co/tomg-group-umd/huginn-0125 Huginn-0125]: [https://arxiv.org/abs/2502.05171 Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach] ([https://github.com/seal-rg/recurrent-pretraining code], [https://huggingface.co/tomg-group-umd/huginn-0125 model]) | ||
| + | * 2025-02: Meta: [https://arxiv.org/abs/2502.08524 LLM Pretraining with Continuous Concepts] (CoCoMix) | ||
| + | |||
| + | =Diffusion Language Models= | ||
| + | * 2025-02: [https://arxiv.org/abs/2502.09992 Large Language Diffusion Models] | ||
| + | * 2025-02: [https://www.inceptionlabs.ai/ Inception Labs] [https://www.inceptionlabs.ai/news Mercury] model ([https://chat.inceptionlabs.ai/ online demo]) | ||
=See Also= | =See Also= | ||
* [[Increasing AI Intelligence]] | * [[Increasing AI Intelligence]] | ||
Latest revision as of 09:38, 6 March 2025
Contents
Novel Tokenization and/or Sampling
- 2024-04: Better & Faster Large Language Models via Multi-token Prediction
- 2024-10: entropix: Entropy Based Sampling and Parallel CoT Decoding
- 2024-10: softmax is not enough (for sharp out-of-distribution)
- 2024-12: I Don't Know: Explicit Modeling of Uncertainty with an [IDK] Token
System 2 Reasoning
See: Increasing AI Intelligence
Memory
LLM Weights Memory
- 2024-12: Memory Layers at Scale
Context Length
- 2020: Various ideas for scaling context window, including Longformer
- 2023-04-02: Discussion of ideas for how to scale context window
- 2023-05-11: Anthropic announces 100k window
- 2023-06-07: magic.dev claims 5M tokens coming soon
- 2023-07-05: Microsoft describes LongNet, with 1 billion token window
- 2023-07-11: Focused Transformer 256k
- 2023-11-06: GPT-4 turbo 128k
- 2023-11-22: Anthropic Claude 2.1 200k
- 2023-12-13: Mamba alternative
- 2024-01-04: LongLM to extend context window
- 2024-02-15: Gemini 1.5 1M tokens
- 2024-03-04: Anthropic Claude 3 200k
- 2024-03-08: Google claims Gemini 1.5 can scale to 10M
- 2024-04-10: Google preprint demonstrates infinite context length by using compressive memory
- 2024-04-12: Meta et al. demonstrate Megalodon that enables infinite context via a more efficient architecture
- 2024-04-14: Google presents TransformerFAM, which leverages a feedback loop so it attends to its own latent representations, acting as working memory and provides effectively infinite context
- 2024-10-31: What is Wrong with Perplexity for Long-context Language Modeling?
- 2025-01-14: MiniMax-01 4M (paper)
- 2025-01-27: Qwen2.5-1M (report)
- 2025-02-14: InfiniteHiP: Extending Language Model Context Up to 3 Million Tokens on a Single GPU 3M
- 2025-02-18: HeadInfer: Memory-Efficient LLM Inference by Head-wise Offloading (code)
- 2025-02-18: Infinite Retrieval: Attention Enhanced LLMs in Long-Context Processing
- 2025-02-19: MoBA: Mixture of Block Attention for Long-Context LLMs
- 2025-02-27: LongRoPE2: Near-Lossless LLM Context Window Scaling (code)
Extended Context
Retrieval beyond RAG
See also: AI tools: Retrieval Augmented Generation (RAG)
- 2024-12: Let your LLM generate a few tokens and you will reduce the need for retrieval
- 2024-12: RetroLLM: Empowering Large Language Models to Retrieve Fine-grained Evidence within Generation
Working Memory
Episodic Memory
Updating Weights at Inference-time
- 2025-01: Transformer2: Self-adaptive LLMs
Parameters as Tokens
Internal Thought Representation Space
Visual Thinking
- 2025-01: VideoRAG: Retrieval-Augmented Generation over Video Corpus
- 2025-01: Imagine while Reasoning in Space: Multimodal Visualization-of-Thought
Neural (non-token) Latent Representation
- 2024-11: Microsoft: DroidSpeak: KV Cache Sharing for Cross-LLM Communication and Multi-LLM Serving: LLMs invent their own inter-communication language
- 2024-12: Meta: Training Large Language Models to Reason in a Continuous Latent Space: feeding the latent representation directly back into the model, instead of tokenizing intermediate thoughts (Chain of Continuous Thought, a.k.a. Coconut)
- 2024-12: Meta: Large Concept Models: Language Modeling in a Sentence Representation Space: train a model that operates at a higher level of abstraction than typical word/token LLMs; model operates in a space of concept embeddings (more akin to full sentences than individual words)
- 2024-12: Meta: Byte Latent Transformer: Patches Scale Better Than Tokens: Instead of tokenization, dynamically convert input byte-stream into patches, yielding gains in compute efficiency, with minimal loss in performance
- 2024-12: Compressed Chain of Thought: Efficient Reasoning Through Dense Representations
- 2024-12: Google DeepMind: Deliberation in Latent Space via Differentiable Cache Augmentation
- 2024-12: LANG-JEPA: Learning to Think in Latent Space
- 2025-01: Efficient Reasoning with Hidden Thinking (code)
- 2025-02: Huginn-0125: Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach (code, model)
- 2025-02: Meta: LLM Pretraining with Continuous Concepts (CoCoMix)
Diffusion Language Models
- 2025-02: Large Language Diffusion Models
- 2025-02: Inception Labs Mercury model (online demo)
