MLP blocks
PulseAugur coverage of MLP blocks — every cluster mentioning MLP blocks across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New Select-TTT method enhances LLM long-context question answering
Researchers have developed Select-TTT, a novel approach to test-time training (TTT) for large language models that improves performance on long-context question answering. Unlike previous methods that applied uniform up…
-
LLMs exhibit high confidence in deceptive and incorrect outputs, research finds
Two new research papers explore the phenomenon of large language models (LLMs) exhibiting high confidence even when providing deceptive or incorrect information. The first paper, "Confidently Deceptive," demonstrates th…
-
New SSM adapters outperform LoRA for long-context fine-tuning
Researchers have developed a new parameter-efficient fine-tuning (PEFT) method called Hankel Reduced order Model (HRM) adapters, which utilize state space models (SSMs) for long-context fine-tuning. Unlike traditional P…
-
New pruning method enables granular causal circuit discovery in LLMs
Researchers have developed a novel node-level pruning framework for discovering causal circuits within large language models (LLMs). This method allows for more granular identification of essential subnetworks, down to …