OpenAI shares more details about its agreement with the Pentagon
By CEO Sam Altman’s own admission, OpenAI’s deal with the Department of Defense was “definitely rushed,” and “the optics don’t look good.”
Like
0
Liked
Liked
Value Bonuses using Ensemble Errors for Exploration in Reinforcement Learning
arXiv:2602.12375v1 Announce Type: new Abstract: Optimistic value estimates provide one mechanism for directed exploration in reinforcement learning (RL). The agent acts greedily with respect to an estimate of the value plus what can be seen as a value bonus. The value bonus can be learned by estimating a value function on reward bonuses, propagating local uncertainties around rewards. However, this approach only increases the value bonus for an action retroactively, after seeing a higher reward bonus from that […]
Like
0
Liked
Liked
Zero-Sacrifice Persistent-Robustness Adversarial Defense for Pre-Trained Encoders
arXiv:2602.11204v1 Announce Type: new Abstract: The widespread use of publicly available pre-trained encoders from self-supervised learning (SSL) has exposed a critical vulnerability: their susceptibility to downstream-agnostic adversarial examples (DAEs), which are crafted without knowledge of the downstream tasks but capable of misleading downstream models. While several defense methods have been explored recently, they rely primarily on task-specific adversarial fine-tuning, which inevitably limits generalizability and causes catastrophic forgetting and deteriorates benign performance. Different with previous works, we propose a […]
Like
0
Liked
Liked
XProvence: Zero-Cost Multilingual Context Pruning for Retrieval-Augmented Generation
arXiv:2601.18886v1 Announce Type: new Abstract: This paper introduces XProvence, a multilingual zero-cost context pruning model for retrieval-augmented generation (RAG), trained on 16 languages and supporting 100+ languages through effective cross-lingual transfer. Motivated by the growing use of RAG systems across diverse languages, we explore several strategies to generalize the Provence framework-which first integrated efficient zero-cost context pruning directly into the re-ranking model-beyond English. Across four multilingual question answering benchmarks, we show how XProvence can prune RAG contexts with […]
Like
0
Liked
Liked
LoPace: A Lossless Optimized Prompt Accurate Compression Engine for Large Language Model Applications
arXiv:2602.13266v1 Announce Type: new Abstract: Large Language Models (LLMs) have changed the way natural language processing works, but it is still hard to store and manage prompts efficiently in production environments. This paper presents LoPace (Lossless Optimized Prompt Accurate Compression Engine), a novel compression framework designed specifically for prompt storage in LLM applications. LoPace uses three different ways to compress data: Zstandard-based compression, Byte-Pair Encoding (BPE) tokenization with binary packing, and a hybrid method that combines the two. […]
Like
0
Liked
Liked
AVP-Pro: An Adaptive Multi-Modal Fusion and Contrastive Learning Approach for Comprehensive Two-Stage Antiviral Peptide Identification
The accurate identification of antiviral peptides (AVPs) is crucial for novel drug development. However, existing methods still have limitations in capturing complex sequence dependencies and distinguishing confusing samples with high similarity. To address these challenges, we propose AVP-Pro, a novel two-stage predictive framework that integrates adaptive feature fusion and contrastive learning. To comprehensively capture the physicochemical properties and deep-seated patterns of peptide sequences, we constructed a panoramic feature space encompassing 10 distinct descriptors and designed a hierarchical fusion […]
Like
0
Liked
Liked
[D] What is your main gripe about ML environments like Colab?
I’ve used Colab a lot over the years and like how easy it is to spin something up. But once I have a few notebooks going, or I try to do anything slightly more serious, it starts feeling messy. I lose track of what’s where, sometimes the runtime dies, and I end up just SSHing into a VM and using VSCode anyway. Maybe I’m just using it wrong. Curious what other people find annoying about these setups. submitted […]
Like
0
Liked
Liked
Active Sequential Learning with Tactile Feedback
We consider the problem of tactile discrimination, with the goal of estimating an underlying state parameter in a sequential setting. If the data is continuous and high-dimensional, collecting enough representative data samples becomes difficult. We present a framework that uses active learning to help with the sequential gathering of data samples, using information-theoretic criteria to find optimal actions at each time step. We consider two approaches to recursively update the state parameter belief: an analytical Gaussian approximation and […]
Like
0
Liked
Liked
A Dynamic Retrieval-Augmented Generation System with Selective Memory and Remembrance
arXiv:2601.02428v1 Announce Type: new Abstract: We introduce emph{Adaptive RAG Memory} (ARM), a retrieval-augmented generation (RAG) framework that replaces a static vector index with a emph{dynamic} memory substrate governed by selective remembrance and decay. Frequently retrieved items are consolidated and protected from forgetting, while rarely used items gradually decay, inspired by cognitive consolidation and forgetting principles. On a lightweight retrieval benchmark, ARM reaches near state-of-the-art performance (e.g., NDCG@5 $approx$ 0.940, Recall@5 $=1.000$) with only $sim$22M parameters in the embedding […]
Like
0
Liked
Liked
The Architectural Shift: How AI Tooling is Decomposing the SaaS Development Stack
Author(s): Shashwata Bhattacharjee Originally published on Towards AI. The narrative of solo founders building eight-figure SaaS businesses using AI tools has become increasingly prevalent in entrepreneurial discourse. While the surface-level story focuses on individual success, the underlying technical transformation represents something far more fundamental: a complete decomposition of the traditional software development stack, enabled by the convergence of large language models, code generation capabilities, and automation frameworks. This analysis examines the technical architecture, economic implications, and systemic changes […]
Like
0
Liked
Liked