SanDisk proposes high-bandwidth flash memory for AI
In a publication dated October 12, 2025, SanDisk proposed High Bandwidth Flash, a NAND design aiming for far greater capacity than HBM with similar read bandwidth and price.
Read articleKnowledge you can apply
Radar
Daily bookmarks: relevant news, facts and discoveries.
SwiReasoning switches reasoning modes
A publication dated October 12, 2025 describes a method that switches between latent and explicit reasoning and reports token-efficiency gains on Qwen3-8B models.
Read articleBPI-AIM7: an AI module with RK3588
Published on October 11, 2025, the item describes the BPI-AIM7 as a low-power AI module based on RK3588 and compatible with the Jetson Nano ecosystem.
Read articleTUMIX coordinates agents to refine responses
Published on October 11, 2025, the post presents TUMIX: agents with different tools solve problems independently, share responses, and refine them over several rounds.
Read articleAgent learning without reward signals
Published on October 10, 2025, the post describes using future states from an agent’s own interactions as supervision, without reward signals, and examines two strategies: implicit world modeling and self-reflection.
Read articleAHN proposes continuous memory for LLMs
A publication dated October 9, 2025, reports that ByteDance introduced Artificial Hippocampus Networks (AHN), an architecture for long-context LLMs that continuously compresses information outside the context window.
Read articleStudy links attention sinks and compression valleys in LLMs
Experiments on models ranging from 410 million to 120 billion parameters associate extreme activation norms of the beginning-of-sequence token in intermediate layers with compression valleys and attention sinks.
Read articleA theory proposes three stages for information flow in LLMs
A publication dated October 9, 2025 links massive activations, attention sinks, and compression valleys, and proposes three stages for information flow in language models.
Read articleJEPA-SCORE estimates density using JEPA encoders
A study published on October 8, 2025 reports that an anti-collapse term in JEPAs can implicitly estimate data density and presents JEPA-SCORE, which does so without retraining.
Read articleUserRL trains agents with simulated users
Published on October 8, 2025, Salesforce’s post presents UserRL, a framework for training agents with standardized environments and simulated users.
Read articleCCA reduces the space used by attention
Published on October 7, 2025, the post describes Compressed Contextual Attention (CCA), which computes attention in a smaller latent space and applies RoPE there.
Read articleCCA proposes compressed attention for long contexts
Published on October 7, 2025, the article presents Compressed Convolutional Attention (CCA), which projects queries, keys, and values to smaller dimensions to perform attention in a compressed latent space.
Read articleFireRedChat explores full-duplex voice interaction
Published on October 7, 2025, this bookmark describes an open-source modular voice interaction system, with cascade and semi-cascade implementations and features for real-time conversation.
Read articleGPU MODE analyzes the AF3 trimul kernel competition
A GPU MODE analysis covers the optimization challenges of the AF3 trimul kernel, how the competition was planned, and what participants implemented.
Read articleCyberBlue bundles blue-team tools in containers
Published on October 6, 2025, the entry describes CyberBlue as a container-based platform bundling open-source tools for SIEM, DFIR, CTI, SOAR, and network analysis.
Read articleIndex catalogs public proof-of-concept exploits for CVEs
Published on October 6, 2025, the briefing describes a GitHub project indexing public proof-of-concept exploits from five sources to support vulnerability research and security testing.
Read articleModel explores neural dynamics and synchronization
Published on October 6, 2025, the article presents the Continuous Thought Machine, which uses neuron-level processing and synchronization as core representations inspired by biological neural dynamics.
Read articleMoE-CL for continual instruction tuning
A paper published on October 6, 2025 proposes the MoE-CL framework for continual instruction tuning and reports evaluations on the MTL5 and Tencent3 benchmarks, plus an A/B test on Tencent. The available summary gives no quantitative results.
Read articleStudy compares in-context and in-weight learning
Published on October 5, 2025, the study reports that a model trained on 12,000 small tasks combined new ideas and found a balance between flexibility and memory.
Read articleGEPA and detecting AI-generated malicious code
A publication dated October 5, 2025 reports that prompts found with GEPA detected malicious behavior in AI-generated code, blocking 90% of malicious submissions with 1% of the audit budget.
Read articleRota Nacional
Bring privacy into your workflow.
30 days, no card, with a starting quota. After that, Pix credit from R$ 10,00.