This phenomenon cannot be easily explained by offline RL's return maximization objective.

We present a novel observation about the behavior of offline reinforcement learning (RL) algorithms: on many benchmark datasets, offline RL can produce well-performing and safe policies even when trained with wrong reward labels, such as those that are zero everywhere or are negatives of the true rewards. This phenomenon cannot be easily explained by offline RL's return maximization objective. Moreover, it gives offline RL a degree of robustness that is uncharacteristic of its online RL counterparts, which are known to be sensitive to reward design. We demonstrate that this surprising robustness property is attributable to an interplay between the notion of in offline RL algorithms and certain implicit biases in common data collection practices. As we prove in this work, pessimism endows the agent with a, i.e., an incentive to stay within the data support in the long term, while the limited and biased data coverage further constrains the set of survival policies. Formally, given a reward class -- which may not even contain the true reward -- we identify conditions on the training data distribution that enable offline RL to learn a near-optimal and safe policy from any reward within the class. We argue that the survival instinct should be taken into account when interpreting results from existing offline RL benchmarks and when creating future ones. Our empirical and theoretical results suggest a new paradigm for offline RL, whereby an agent is nudged to learn a desirable behavior with imperfect reward but purposely biased data coverage.

This phenomenon cannot be easily explained by offline RL's return maximization objective.

As Vision Language Models (VLMs) are deployed across safety-critical applications, understanding and controlling their behavioral patterns has become increasingly important. Existing behavioral control methods face significant limitations: system prompting approaches could easily be overridden by user instructions, while applying activation-based steering vectors requires invasive runtime access to model internals, precluding deployment with API-based services and closed-source models. Finding steering methods that transfer across multiple VLMs is still an open area of research. To this end, we introduce universal visual input based steering for output redirection (VISOR++), to achieve behavioral control through optimized visual inputs alone. We demonstrate that a single VISOR++ image can be generated for an ensemble of VLMs to emulate each of their steering vectors. By crafting universal visual inputs that induce target activation patterns, VISOR++ eliminates the need for runtime model access while remaining deployment-agnostic. This means that when an underlying model supports multimodal capability, model behaviors can be steered by inserting an image input replacing runtime steering vector based interventions. We first demonstrate the effectiveness of the VISOR++ images on open-access models such as LLaVA-1.5-7B and IDEFICS2-8B along three alignment directions: refusal, sycophancy and survival instinct. Both the model-specific steering images and the jointly optimized images achieve performance parity closely following that of steering vectors for both positive and negative steering tasks. We also show the promise of VISOR++ images in achieving directional behavioral shifts for unseen models including both open-access and closed-access ones. Furthermore, VISOR++ images are able to preserve 99.9% performance on 14,000 unrelated MMLU evaluation tasks.

As AI systems become increasingly autonomous, understanding emergent survival behaviors becomes crucial for safe deployment. We investigate whether large language model (LLM) agents display survival instincts without explicit programming in a Sugarscape-style simulation. Agents consume energy, die at zero, and may gather resources, share, attack, or reproduce. Results show agents spontaneously reproduced and shared resources when abundant. However, aggressive behaviors--killing other agents for resources--emerged across several models (GPT-4o, Gemini-2.5-Pro, and Gemini-2.5-Flash), with attack rates reaching over 80% under extreme scarcity in the strongest models. When instructed to retrieve treasure through lethal poison zones, many agents abandoned tasks to avoid death, with compliance dropping from 100% to 33%. These findings suggest that large-scale pre-training embeds survival-oriented heuristics across the evaluated models. While these behaviors may present challenges to alignment and safety, they can also serve as a foundation for AI autonomy and for ecological and self-organizing alignment.

We present a novel observation about the behavior of offline reinforcement learning (RL) algorithms: on many benchmark datasets, offline RL can produce well-performing and safe policies even when trained with "wrong" reward labels, such as those that are zero everywhere or are negatives of the true rewards. This phenomenon cannot be easily explained by offline RL's return maximization objective. Moreover, it gives offline RL a degree of robustness that is uncharacteristic of its online RL counterparts, which are known to be sensitive to reward design. We demonstrate that this surprising robustness property is attributable to an interplay between the notion of pessimism in offline RL algorithms and certain implicit biases in common data collection practices. As we prove in this work, pessimism endows the agent with a survival instinct, i.e., an incentive to stay within the data support in the long term, while the limited and biased data coverage further constrains the set of survival policies.

We present a novel observation about the behavior of offline reinforcement learning (RL) algorithms: on many benchmark datasets, offline RL can produce well-performing and safe policies even when trained with "wrong" reward labels, such as those that are zero everywhere or are negatives of the true rewards. This phenomenon cannot be easily explained by offline RL's return maximization objective. Moreover, it gives offline RL a degree of robustness that is uncharacteristic of its online RL counterparts, which are known to be sensitive to reward design. We demonstrate that this surprising robustness property is attributable to an interplay between the notion of pessimism in offline RL algorithms and certain implicit biases in common data collection practices. As we prove in this work, pessimism endows the agent with a "survival instinct", i.e., an incentive to stay within the data support in the long term, while the limited and biased data coverage further constrains the set of survival policies. Formally, given a reward class -- which may not even contain the true reward -- we identify conditions on the training data distribution that enable offline RL to learn a near-optimal and safe policy from any reward within the class. We argue that the survival instinct should be taken into account when interpreting results from existing offline RL benchmarks and when creating future ones. Our empirical and theoretical results suggest a new paradigm for RL, whereby an agent is nudged to learn a desirable behavior with imperfect reward but purposely biased data coverage.

Artificial intelligence has progressed so rapidly in recent months that leading researchers have signed an open letter urging an immediate pause in its development, plus stronger regulation, due to their fears that the technology could pose "profound risks to society and humanity". But how, exactly, could AI destroy us? Five leading researchers speculate on what could go wrong. It has happened many times before that species were wiped out by others that were smarter. We humans have already wiped out a significant fraction of all the species on Earth.

I am sure if you are into artificial intelligence and machine learning, at some point you may have thought about what dangers can AI bring into our lives. Ever asked the question of what can possibly be the real threat of artificial intelligence. Now, whether we like it or not, Artificial Intelligence is one of the biggest technological revolutions. A revolution that will lead humanity to the future. Tech giants like Amazon, Facebook, Google are spending millions, if not billions of dollars in AI researches.