Meet Anca Mihalache, a person who believes real impact lives where purpose meets meaningfulness. With a sharp eye for details and a healthy skepticism,
Stories
Get to know AscentCore through the eyes of our talented team members.
Discover the story of Horia Balc, a Software Engineer who brings the same determination to code as he does to the starting line of a
Discover the story of Laura Marinoiu, an Engineering Manager who sees leadership as an ongoing journey shaped by curiosity, trust, and strong teams. Guided by
At AscentCore, our mentoring program creates space for employees to learn from one another, share perspectives, and improve the way they work. These relationships build
Amalia Drăguș: A Journey of Leadership and Growth from Engineer to Associate Director in Engineering
Discover the inspiring journey of Amalia Drăguș, one of our leaders who balances professional life with personal growth. From leading distributed teams to embracing creativity
In this edition of our Employee Stories, we reconnect with Corina Velea, one of AscentCore’s talented Technical Leads. Corina’s journey reflects an inspiring evolution from
An organisation that cannot remember what it decided, or why, is condemned to decide the same things over and over again, each time believing it is the first.
Large language models respond to emotionally charged inputs with contextually appropriate outputs, but the mechanism by which they represent, propagate, and modulate emotional tone through their internal layers remains poorly understood. Do emotions "live" in specific layers? Is the signal carried by the attention mechanism, the MLP, or the residual stream itself? And when a model is instructed to be a "helpful assistant," does its internal representation remain emotionally neutral, or does it mirror the user's emotional state?
The tool is never the bottleneck. The bottleneck is everything the tool cannot see, cannot access, and does not know it should ask about. The
An organisation that cannot remember what it decided, or why, is condemned to decide the same things over and over again, each time believing it is the first.
This report presents the results of a systematic evaluation of 22 quantized open-source language models across description generation tasks, measuring quality, JSON reliability, and inference efficiency.
Your best feature may be destroying your margins, and your engineering team has no idea. This article isn’t about AI as a productivity tool.
An organisation that cannot remember what it decided, or why, is condemned to decide the same things over and over again, each time believing it is the first.
Large language models respond to emotionally charged inputs with contextually appropriate outputs, but the mechanism by which they represent, propagate, and modulate emotional tone through their internal layers remains poorly understood. Do emotions "live" in specific layers? Is the signal carried by the attention mechanism, the MLP, or the residual stream itself? And when a model is instructed to be a "helpful assistant," does its internal representation remain emotionally neutral, or does it mirror the user's emotional state?
The tool is never the bottleneck. The bottleneck is everything the tool cannot see, cannot access, and does not know it should ask about. The
An organisation that cannot remember what it decided, or why, is condemned to decide the same things over and over again, each time believing it is the first.
This report presents the results of a systematic evaluation of 22 quantized open-source language models across description generation tasks, measuring quality, JSON reliability, and inference efficiency.
Your best feature may be destroying your margins, and your engineering team has no idea. This article isn’t about AI as a productivity tool.