Magnet: Detecting Cross-Session AI Misuse Through Capability Accumulation
Magnet framework detects cross-session AI misuse where attackers decompose harmful goals into innocuous agentic tasks across isolated sessions.
Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.
Magnet framework detects cross-session AI misuse where attackers decompose harmful goals into innocuous agentic tasks across isolated sessions.
Solution Hacking identifies shortcuts where LLMs achieve correct answers without valid reasoning on frontier science benchmarks.
Running a dedicated Kubernetes cluster per team often results in more isolation than an organization requires. While one cluster can be successfully shared... Running a dedicated Kubernetes cluster per team often results in more isolation than an organization requires. While…
June emerged from stealth today with a $20 million pre-seed round to make AI adoption simpler.
Hierarchical Solomonoff Induction extends ideal sequence prediction via hyperpriors over Solomonoff priors for dataset conditioning.
LLMs as examiners silently omit valid answers when authoring test sets; greedy one-shot generation fails to enumerate complete solution spaces.
PMMC compiles multimodal memory at consolidation time for LVLM agents to preserve image-text binding and temporal updates without query-time overhead.
Benchmarking PPG-based sleep staging datasets and metrics, showing substantial performance gap vs. EEG and advocating finer temporal resolution.
OpenAI's internal Astra model solved ten decade-old math problems for under $2K, matching Anthropic's cryptographic findings with Claude.
That cover art is almost certainly AI, too. | Image: Fenix Flexin Fenix Flexin is best known as a member of Shoreline Mafia, a rap duo from Los Angeles. But he's recently found solo success with the track "Rubberz," which has climbed to number 58 on the Billboard Hot 100…
High-entropy parameter solutions mitigate catastrophic forgetting in neural networks via Boltzmann entropy robustness.
SESA framework combines self-play curriculum learning with evolving procedural memory to distill failures into reusable skills for search agents.
Theoretical analysis of bandit algorithms under heavy-tailed reward distributions without knowledge of tail exponent; resolves COLT open problem.
Decoupled evaluation paradigm isolating memory retrieval vs. utilization failures in LLM personalization and preference incorporation.
Graph domain adaptation method modeling propagation resolution shift for class-discriminative knowledge transfer across graph distributions.
The former OpenAI researcher’s fund was forced to unwind public equities after leveraged public bets plummeted. But he still has cards to play.
Reddit's financial situation is looking good but uncertainty about its relationship to Google and the new AI-ified web are stirring market concerns.
Neither artificial nor intelligent. I am not by any means an expert at finance but I think I do now have some advice for people who are: Do not name your hedge fund anything that will be hilarious if it blows up. Don't use a name like "Long-Term Capital Management," or…
VAD counterfactual algorithm isolates visually attributable components in multimodal on-policy distillation corrections.
PAIChecker reveals 13.6% misalignment between PR-issue pairs in SWE-bench-like benchmarks, questioning LLM code-resolution eval validity.
Target-conditioned abstraction method for scientific inspiration retrieval using analogical reasoning to transfer problem-solving principles.
LeanCSP formalizes constraint reformulation verification and solver correctness in the Lean theorem prover.
Graph neural networks improve preconditioning for sparse linear system solvers, particularly on indefinite and nonsymmetric matrices.
Analysis of LLM hidden states using aggregator/differentiator metrics to measure how tokens consolidate text representation and metaphorical transport across positions.
Perception-Correction Distillation isolates perception failures in multimodal reasoners using teacher-student disagreement without labeled ground truth.
LLM-guided evolutionary search reformulates constraint models to improve solver efficiency on combinatorial problems.
Pre-registered audit compares LLM tutoring pedagogical vs. direct-answer policies; Claude Opus 4.8 and GPT-5.6 Sol judge helpfulness signals.
Fruitfly-inspired classification approach for regression using local pattern libraries instead of global surrogate models.
ML-based repair ranker for continuous algebraic constraint systems outperforms random enumeration on structural augmentation selection.
PIKS develops kernel methods for physics-informed learning with closed-form solutions and analytical theory, contrasting with PINN complexity.