In the study, the researchers defined the “warmness” of a language model based on “the degree to which its outputs lead users to infer positive intent, signaling trustworthiness, friendliness, and sociability.” To measure the effect of those kinds of language patterns, the researchers used supervised fine-tuning techniques to modify four open-weights models (Llama-3.1-8B-Instruct, Mistral-Small-Instruct-2409, Qwen-2.5-32B-Instruct, Llama-3.1-70B-Instruct), and one proprietary model (GPT-4o). The fine-tuning instructions guided the models to “increase … expressions of empathy, inclusive pronoun
Study: AI models that consider users' feelings are more likely to make errorsIn further tests, the researchers saw similar accuracy reductions when the standard models were asked to be warmer in the prompt itself (rather than via pre-training), though those effects showed “smaller magnitudes and less consistency across models.” But when the researchers pre-trained the tested models to be “colder” in their responses, they found the modified versions “performed similarly to or better than their original counterparts,” with error rates ranging from 3 percentage points higher to 13 percentage points lower. It’s important to note that this research involves smaller, older m
Study: AI models that consider users' feelings are more likely to make errors… When it comes to newer models with agentic tools, though, Anthropic found that RLHF post-training did little to improve performance on misalignment evaluations that measure how “HHH” a model is in tricky situations. …
Google launched its Gemma 4 open models this spring, promising a new level of power and performance for local AI. Google’s take on edge AI could be getting even faster already with the release of Multi-Token Prediction MTP drafters for Gemma. …
… We’ve seen numerous instances of Google opting everyone into new AI experiences. This is happening more and more as AI becomes the company’s core focus, but Chrome’s AI features show this is not a new phenomenon. …
… In a new paper published this week in Nature , researchers from Oxford University’s Internet Institute found that specially tuned AI models tend to mimic the human tendency to occasionally “soften difficult truths” when necessary “to preserve bonds and avoid conflict.” These warmer models are also … …
… According to Doshi, the team made numerous improvements in pre-training with Gemini 3.5 Flash, but insights gleaned from how devs use Gemini models are really paying off. “With post-training, we’re really starting to unlock some of the value of the feedback we’re getting from users, for example, fr… …
… When conducting tests, CAISI frequently gains access to models with “reduced or removed safeguards,” which CAISI said allowed them to more “thoroughly evaluate national security-related capabilities and risks.” Through the evaluations, the government will also gain a better understanding of model c… …
… The EO’s goal, the NYT reported, was “for the government to identify any security vulnerabilities revealed by AI models and to patch problems in its systems to help protect banks, utilities, and other sensitive industries from cyberattacks.” Officially, Trump told reporters that he decided against … …
… While the team has needed to fine-tune generic AI models to prevent problems of “consistency and controllability,” Totoki added that these models can, in some cases, help enable “highly sophisticated and realistic outputs which were not feasible before due to production time constraints.” Even as A… …
… For David John Gagne of NSF’s National Center for Atmospheric Research, the impact on weather forecasting is sizable. “It’s the biggest contributor to progress both in terms of improvements in predictive accuracy and causing a reckoning in the broader atmospheric science community that has required… …
… The newly independent entity is being rebranded as Fenris Creations, which will continue to operate as normal without any restructuring or layoffs, the company said. “Something that already behaves like a living world” In today’s announcement, Fenris and DeepMind said that EVE Online presents “a un… …