Llama 2 7B
Emerging6papers using it
2025first seen
'LLaMA2-7B' is a large language model benchmark used to evaluate the performance and efficiency of compression schemes like Tensor Mixture (MixT) in maintaining accuracy while reducing model parameters and computational overhead.
Papers using Llama 2 7B (6)
- SoLA: Leveraging Soft Activation Sparsity and Low-Rank Decomposition for Large Language Model CompressionA general tensor-structured compression scheme for efficient large language modelsAnchors in the Machine: Behavioral and Attributional Evidence of Anchoring Bias in LLMsDarwinLM: Evolutionary Structured Pruning of Large Language ModelsFairy2i: Training Complex LLMs from Real LLMs with All Parameters in {pm 1, pm i}RoRA: Efficient Fine-Tuning of LLM with Reliability Optimization for
Rank Adaptation