Cisco vs Palo Alto Networks

Small Language Models: When Smaller Is Better

This MLPills newsletter issue explains small language models (SLMs) — compact AI models designed to run under resource constraints such as limited memory, power, and latency budgets. It clarifies that 'small' is a comparative concept rather than a specific parameter count, and distinguishes SLMs from quantized frontier models and distillation. The article covers four main routes to building SLMs: curated data training, distillation, pruning, and quantization, and highlights examples including Microsoft's Phi-4 family, Google's Gemma 3n and Gemma 4 edge models, Hugging Face's SmolLM3, and Cisco's Antares vulnerability-localization models. It describes ideal use cases like classification, entity extraction, and tool selection, and recommends a layered architecture using deterministic code, small models, large models, and human oversight. The piece also cautions about evaluation, over-pruning, and privacy limitations of local inference.

• A 4B parameter model at 4-bit precision needs roughly 2 GB of memory; the same model at 16-bit needs roughly 8 GB.
• Microsoft's Phi-4 is a 14B parameter model, with a 3.8B Phi-4-Mini sibling and a Phi-4-Multimodal variant.
• Google's Gemma 3n models used per-layer embeddings, KV cache sharing, and activation quantization to reduce memory footprints.

Defining the Standard for AI Security

Palo Alto Networks announces a new AI security initiative, with Anand Oswal discussing the company's approach to securing AI coding and other multidimensional AI security challenges.

Palo Alto Networks Completes Acquisition of CyberArk to Secure the AI Era

Palo Alto Networks announces the completion of its acquisition of CyberArk, redefining identity security for the modern enterprise. The press release is available on the company's newsroom.

Cybersecurity stock Zscaler's charts turn bullish: Jay Woods

In a CNBC Pro segment, technician Jay Woods discusses Zscaler (ZS), a cybersecurity company that has underperformed its sector peers. While CrowdStrike, Fortinet, Okta, and Palo Alto Networks have gained over 100% year-to-date, Zscaler shares fell as much as 65% from 52-week highs and remain down 13% this year. However, Woods observes a bottoming formation on both daily and weekly charts, with a bullish crossover in MACD and a breakout above $195 resistance. He suggests upside targets of $250-$265, representing 25-30% potential upside, with support at $165. The analysis highlights relative rotation within the cybersecurity sector, indicating that laggards like Zscaler may see renewed strength.

• Zscaler (ZS) shares fell as much as 65% from 52-week highs and remain down 13% year-to-date.
• CrowdStrike, Fortinet, Okta, and Palo Alto Networks have each gained over 100% year-to-date.
• Jay Woods identifies a bullish double bottom formation on Zscaler's weekly chart.