Phi-3.5-mini
PulseAugur coverage of Phi-3.5-mini — every cluster mentioning Phi-3.5-mini across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New research probes LLM judge subjectivity and evaluation pitfalls · 7 sources tracked
Recent research explores the complexities and potential pitfalls of using Large Language Models (LLMs) as judges for evaluating other AI models and outputs. One study highlights that simply reading the first token of an…
-
New dataset OpenDiscoveryTrace tracks AI scientist reasoning processes
A new dataset called OpenDiscoveryTrace has been released, containing 558 detailed AI scientific agent trajectories. This dataset captures the step-by-step reasoning processes of models, not just their final outputs, to…
-
TriAgent cuts LLM costs for financial sentiment analysis
Researchers have developed TriAgent, a novel multi-agent system designed to reduce the cost of financial sentiment analysis using large language models. The system stratifies agents by contextual granularity, employing …
-
AI benchmarking tools: Custom builds vs. existing solutions explored
The author of a blog focused on AI hardware and model performance investigated whether their custom-built benchmarking tools were necessary or if existing solutions could have been utilized. They found that while hardwa…
-
LlamaGuard fails to stop RAG injection attacks, PromptGuard succeeds
A security researcher found that LlamaGuard-3-1B, a model designed to protect against harmful content, completely failed to detect 10 different RAG injection attacks. These attacks, which have previously succeeded again…
-
Small LLMs exhibit positional bias, not answer avoidance, when sandbagging
New research indicates that smaller language models (7-9 billion parameters) exhibit a positional bias when instructed to "sandbag" or underperform, rather than avoiding correct answers. This bias causes models like Lla…
-
New architecture enables privacy-preserving LLM personalization with deletable user proxies
Researchers have developed a novel three-layer architecture designed to enhance privacy in personalized large language models. This system separates user-specific data from the core model weights by utilizing composable…