7 papers
Quantization Inflates Reasoning: Token Inflation as a Hidden Cost of Low-Bit Reasoning Models
Xinyu Lian, Walid Krichene, Beichen Huang +4
Quantization is widely used to reduce the inference cost of large language models, but its effect on reasoning models is not fully captured by final-answer accuracy or per-token la…
AutoSP: Unlocking Long-Context LLM Training Via Compiler-Based Sequence Parallelism
Ahan Gupta, Zhihao Wang, Neel Dani +3
Large-language-models (LLMs) demonstrate enormous utility in long-context tasks which require processing prompts that consist of tens to hundreds of thousands of tokens. However, e…
MAC-Attention: a Match-Amend-Complete Scheme for Fast and Accurate Attention Computation
Jinghan Yao, Sam Adé Jacobs, Walid Krichene +2
Long-context decoding in LLMs is IO-bound: each token re-reads an ever-growing KV cache. Prior accelerations cut bytes via compression, which lowers fidelity, or selection/eviction…
The Bias is in the Details: An Assessment of Cognitive Bias in LLMs
R. Alexander Knipper, Charles S. Knipper, Kaiqi Zhang +3
As Large Language Models (LLMs) are increasingly embedded in real-world decision-making processes, it becomes crucial to examine the extent to which they exhibit cognitive biases.…
SuperOffload: Unleashing the Power of Large-Scale LLM Training on Superchips
Xinyu Lian, Masahiro Tanaka, Olatunji Ruwase +1
The emergence of Superchips represents a significant advancement in next-generation AI hardware. These Superchips employ a tightly coupled heterogeneous architecture that integrate…
Universal Checkpointing: A Flexible and Efficient Distributed Checkpointing System for Large-Scale DNN Training with Reconfigurable Parallelis
Xinyu Lian, Sam Ade Jacobs, Lev Kurilenko +4
Deep neural network (DNN) training continues to scale rapidly in terms of model size, data volume, and sequence length, to the point where multiple machines are required to fit lar…