Can Pakistan join the Gaza stabilisation force? Two Palestinian children have been killed in an Israeli drone attack on the town of Bani Suheila, east of Khan Younis in the southern Gaza Strip, according to medics. Witnesses told Al Jazeera that drones dropped a bomb on a group of civilians near al-Farabi School on Saturday morning, killing two brothers, Juma and Fadi Tamer Abu Assi. The area targeted by the attack lies beyond the so-called Yellow Line marking the Israeli forces' redeployment boundary agreed under the ceasefire agreement, which took effect on October 10, sources said. Earlier on Saturday, the Israeli military launched ground, naval, and air attacks on several parts of Gaza.

Resource constraints can fundamentally change both learning and decision-making. We explore how memory constraints influence an agent's performance when navigating unknown environments using standard reinforcement learning algorithms. Specifically, memory-constrained agents face a dilemma: how much of their limited memory should be allocated to each of the agent's internal processes, such as estimating a world model, as opposed to forming a plan using that model? We study this dilemma in MCTS- and DQN-based algorithms and examine how different allocations of memory impact performance in episodic and continual learning settings.

This study reveals an unexpected parallel between instructible vision-language models (VLMs) and human cognitive disorders, specifically constructive apraxia. We tested 25 state-of-the-art VLMs, including GPT-4 Vision, DALL-E 3, and Midjourney v5, on their ability to generate images of the Ponzo illusion, a task that requires basic spatial reasoning and is often used in clinical assessments of constructive apraxia. Remarkably, 24 out of 25 models failed to correctly render two horizontal lines against a perspective background, mirroring the deficits seen in patients with parietal lobe damage. The models consistently misinterpreted spatial instructions, producing tilted or misaligned lines that followed the perspective of the background rather than remaining horizontal. This behavior is strikingly similar to how apraxia patients struggle to copy or construct simple figures despite intact visual perception and motor skills. Our findings suggest that current VLMs, despite their advanced capabilities in other domains, lack fundamental spatial reasoning abilities akin to those impaired in constructive apraxia. This limitation in AI systems provides a novel computational model for studying spatial cognition deficits and highlights a critical area for improvement in VLM architecture and training methodologies.

We explore colour versus shape goal misgeneralization originally demonstrated by Di Langosco et al. (2022) in the Procgen Maze environment, where, given an ambiguous choice, the agents seem to prefer generalization based on colour rather than shape. After training over 1,000 agents in a simplified version of the environment and evaluating them on over 10 million episodes, we conclude that the behaviour can be attributed to the agents learning to detect the goal object through a specific colour channel. This choice is arbitrary. Additionally, we show how, due to underspecification, the preferences can change when retraining the agents using exactly the same procedure except for using a different random seed for the training run. Finally, we demonstrate the existence of outliers in out-of-distribution behaviour based on training random seed alone.

In another article, we touched a bit on generalization. What is the relationship between the generalization error and the training error? Generalization is the concept of the machine learning algorithm being able to produce good predictions on previously unseen inputs. The red line represents the training error. If the horizontal axis is the quantity of training examples or time, depending on how you like to think about it, then with time this training error gets smaller and smaller.

In a previous document I described how bayesian models can recursively update, thus making them ideal as a starting point for designing streaming machine learning models. In this document I will describe a different method proposed by Crammer et al. which includes a passive agressive approach to model updates. I will focus on intuition first before moving to the mathy bits. Let's say you're doing a regression for a single data point $d_i$. If you only have one datapoint then you don't know what line is best.

Identifying lanes on the road is a common task performed by all human drivers to ensure their vehicles are within lane constraints when driving, so as to make sure traffic is smooth and minimise chances of collisions with other cars in nearby lanes. Similarly, it is a critical task for an autonomous vehicle to perform. It turns out that recognising lane markings on roads is possible using well known computer vision techniques. We will cover how to use various techniques to identify and draw the inside of a lane, compute lane curvature, and even estimate the vehicle's position relative to the center of the lane. The first step we will take is to find the calibration matrix, along with distortion coefficients for the camera that was used to take pictures of the road.