Ishmam Zabir

Abdelrahman Abouelenin, Atabak Ashfaq, Adam Atkinson et al. · microsoft-research

We introduce Phi-4-Mini and Phi-4-Multimodal, compact yet highly capable language and multimodal models. Phi-4-Mini is a 3.8-billion-parameter language model trained on high-quality web and synthetic data, significantly outperforming recent open-source models of similar size and matching the performance of models twice its size on math and coding tasks requiring complex reasoning. This achievement is driven by a carefully curated synthetic data recipe emphasizing high-quality math and coding datasets. Compared to its predecessor, Phi-3.5-Mini, Phi-4-Mini features an expanded vocabulary size of 200K tokens to better support multilingual applications, as well as group query attention for more efficient long-sequence generation. Phi-4-Multimodal is a multimodal model that integrates text, vision, and speech/audio input modalities into a single model. Its novel modality extension approach leverages LoRA adapters and modality-specific routers to allow multiple inference modes combining various modalities without interference. For example, it now ranks first in the OpenASR leaderboard to date, although the LoRA component of the speech/audio modality has just 460 million parameters. Phi-4-Multimodal supports scenarios involving (vision + language), (vision + speech), and (speech/audio) inputs, outperforming larger vision-language and speech-language models on a wide range of tasks. Additionally, we experiment to further train Phi-4-Mini to enhance its reasoning capabilities. Despite its compact 3.8-billion-parameter size, this experimental version achieves reasoning performance on par with or surpassing significantly larger models, including DeepSeek-R1-Distill-Qwen-7B and DeepSeek-R1-Distill-Llama-8B.

Batuhan K. Karaman, Ishmam Zabir, Alon Benhaim et al.

Achieving both high safety and high usefulness simultaneously in large language models has become a critical challenge in recent years.Models often exhibit unsafe behavior or adopt an overly cautious approach leading to frequent overrefusal of benign prompts, which reduces their usefulness. A major factor underlying these behaviors is how the models are finetuned and aligned, particularly the nature and extent of the data used.In this work, we examine how overgenerating finetuning data with advanced teacher models (e.g., GPT-4o)-covering both general-purpose and toxic prompts-affects safety and usefulness in instruction-following language models.Additionally, we present POROver, an alignment strategy designed for models that are highly safe but prone to overrefusal. POROver employs preference optimization algorithms and leverages completions from an advanced teacher model to reduce overrefusals while maintaining safety.Our results show that overgenerating completions for general-purpose prompts significantly boosts safety with only a minimal impact on usefulness. Specifically, the F1 score calculated between safety and usefulness increases from 74.4% to 91.8% because of a substantial rise in safety. Moreover, overgeneration for toxic prompts raises usefulness from 11.1% to 57.6% while preserving safety. Finally, applying POROVer increases usefulness further-from 57.6% to 82.1%-while keeping safety at comparable levels. Our data and code are available at https://github.com/batuhankmkaraman/POROver.

Ishmam Zabir, Evangelos E. Papalexakis

The PARAFAC tensor decomposition has enjoyed an increasing success in exploratory multi-aspect data mining scenarios. A major challenge remains the estimation of the number of latent factors (i.e., the rank) of the decomposition, which yields high-quality, interpretable results. Previously, we have proposed an automated tensor mining method which leverages a well-known quality heuristic from the field of Chemometrics, the Core Consistency Diagnostic (CORCONDIA), in order to automatically determine the rank for the PARAFAC decomposition. In this work we set out to explore the trade-off between 1) the interpretability/quality of the results (as expressed by CORCONDIA), and 2) the predictive accuracy of the results, in order to further improve the rank estimation quality. Our preliminary results indicate that striking a good balance in that trade-off benefits rank estimation.