Discovering achievements with a hierarchical structure in procedurally generated environments presents a significant challenge. This requires an agent to possess a broad range of abilities, including generalization and long-term reasoning. Many prior methods have been built upon model-based or hierarchical approaches, with the belief that an explicit module for long-term planning would be advantageous for learning hierarchical dependencies. However, these methods demand an excessive number of environment interactions or large model sizes, limiting their practicality. In this work, we demonstrate that proximal policy optimization (PPO), a simple yet versatile model-free algorithm, outperforms previous methods when optimized with recent implementation practices. Moreover, we find that the PPO agent can predict the next achievement to be unlocked to some extent, albeit with limited confidence. Based on this observation, we introduce a novel contrastive learning method, called achievement distillation, which strengthens the agent's ability to predict the next achievement. Our method exhibits a strong capacity for discovering hierarchical achievements and shows state-of-the-art performance on the challenging Crafter environment in a sample-efficient manner while utilizing fewer model parameters.

Is an AI version of Mark Zuckerberg - or any boss - a good plan? Feedback has learned that, according to reports, Meta is building an AI version of Mark Zuckerberg to interact with staff. Feedback hopes this doesn't become a trend Feedback has had a number of bosses over the years. One reorganised the company at which we worked in such a way that our job no longer existed. However, none of them was an AI.

Speech language models (SpeechLMs) process and generate acoustic data only, without textual supervision. In this work, we propose TWIST, a method for training SpeechLMs using a warm-start from a pretrained textual language models. We show using both automatic and human evaluations that TWIST outperforms a cold-start SpeechLM across the board. We empirically analyze the effect of different model design choices such as the speech tokenizer, the pretrained textual model, and the dataset size. We find that model and dataset scale both play an important role in constructing better-performing SpeechLMs. Based on our observations, we present the largest (to the best of our knowledge) SpeechLM both in terms of number of parameters and training data. We additionally introduce two spoken versions of the StoryCloze textual benchmark to further improve model evaluation and advance future research in the field. We make speech samples, code and models publicly available.2

We perform a non-asymptotic analysis of the contrastive divergence (CD) algorithm, a training method for unnormalized models. While prior work has established that (for exponential family distributions) the CD iterates asymptotically converge at an O(n 1/3) rate to the true parameter of the data distribution, we show, under some regularity assumptions, that CD can achieve the parametric rate O(n 1/2). Our analysis provides results for various data batching schemes, including the fully online and minibatch ones. We additionally show that CD can be near-optimal, in the sense that its asymptotic variance is close to the Cramér-Rao lower bound.

The families of victims of a school shooting in a remote Canadian Rockies town are suing artificial intelligence company OpenAI in a United States federal court, alleging that the ChatGPT maker failed to alert police to the shooter's alarming interactions with the chatbot. A lawsuit filed on Wednesday on behalf of 12-year-old Maya Gebala, who was critically injured in the February shooting, is among the first of more than two dozen cases from families in Tumbler Ridge, British Columbia, in what their lawyers say represents "an entire community stepping forward to hold OpenAI accountable". The cases represent the families of the five slain children targeted in the school shooting. Those include Zoey Benoit, Abel Mwansa Jr, Ticaria "Tiki" Lampert, Kylie Smith, all 12, and Ezekiel Schofield, 13, as well as education assistant Shannda Aviugana-Durand. Jesse Van Rootselaar, whose interactions with ChatGPT are at the centre of the lawsuits, shot her mother and stepbrother at home before killing an educational assistant and five students aged 12 to 13 at her former school on February 10, according to police.

May's PlayStation Plus lineup includes Nine Sols, EA FC 26, Wuchang May's PlayStation Plus lineup includes Nine Sols, EA FC 26, Wuchang This coming month of May, PlayStation Plus is getting three pretty great games from three distinct genres: is a 2D action-platformer, is a soccer simulator and is a soulslike RPG. All of these titles will be available to PlayStation Plus subscribers on May 5. You may remember RedCandleGames from the 2019 Chinese censorship campaign targeting the studio's title . It's a dark-fantasy story set during the Ming Dynasty, developed by Chinese studio Leenzee and published by 505 Games. The May PlayStation Plus drop includes the PlayStation Plus Icons Pack.

With US restrictions limiting its access to advanced tech, SenseTime is doubling down on open source with a new model optimized to run on Chinese-made chips. SenseTime, a Chinese AI company best known for its facial recognition technology, released a new open source model on Tuesday that it claims can both generate and interpret images far faster than top models developed by US competitors. SenseNova U1 could help the company reclaim lost ground after it slipped from its place among the leading players in China's AI development race. The model's secret sauce is its ability to "read" images without translating them to text first, speeding up the process and reducing the amount of computing power required. "The model's entire reasoning process is no longer limited to text. It can reason with images as well," Dahua Lin, cofounder and chief scientist at SenseTime, said in an interview with WIRED.

We study the problem of (ϵ,δ)-certified machine unlearning for minimax models. Most of the existing works focus on unlearning from standard statistical learning models that have a single variable and their unlearning steps hinge on the direct Hessian-based conventional Newton update. We develop a new (ϵ,δ)-certified machine unlearning algorithm for minimax models. It proposes a minimax unlearning step consisting of a total Hessian-based complete Newton update and the Gaussian mechanism borrowed from differential privacy. To obtain the unlearning certification, our method injects calibrated Gaussian noises by carefully analyzing the "sensitivity" of the minimax unlearning step (i.e., the closeness between the minimax unlearning variables and the retraining-from-scratch variables).

We study the problem of (ϵ,δ)-certified machine unlearning for minimax models. Most of the existing works focus on unlearning from standard statistical learning models that have a single variable and their unlearning steps hinge on the direct Hessian-based conventional Newton update. We develop a new (ϵ,δ)-certified machine unlearning algorithm for minimax models. It proposes a minimax unlearning step consisting of a total Hessian-based complete Newton update and the Gaussian mechanism borrowed from differential privacy. To obtain the unlearning certification, our method injects calibrated Gaussian noises by carefully analyzing the "sensitivity" of the minimax unlearning step (i.e., the closeness between the minimax unlearning variables and the retraining-from-scratch variables).

When you purchase through links in our articles, we may earn a small commission. If you're using long and overly specific prompts with ChatGPT's latest model, you're doing it wrong. OpenAI's latest and most powerful model, GPT-5.5, has been topping benchmark charts and impressing users with its coding and reasoning abilities, not to mention the sheer quantity of facts at its fingertips. But while ChatGPT's latest model doesn't require the hand-holding that older models did, it also gets fussy with the longer, highly detailed prompts that might have worked well in the past. If you're seeing worse performance with GPT-5.5 than you had with previous models, it might be your prompt constructions.