Engineering ·
CERN slides introduce LLM training
A post dated August 31, 2026, says slides from a CERN talk on LLM training are available as a PDF and that an online recording is available.
Radar
Daily bookmarks: relevant news, facts and discoveries.
Articles 1219
Engineering ·
Article reviews policy gradients and variants
An article on reinforcement learning foundations derives the policy gradient algorithm and discusses variants, providing review material for engineers.
Engineering ·
KV Cache: Queries Are Not Kept
A post dated August 30, 2026 distinguishes the one-time use of queries from the retention of keys and values in the KV cache during inference.
Engineering ·
Hy4: calibration-guided low-bit compression
A post reports reducing the Hy4 preview from 1.5 TB to 200 GB with per-layer precision settings, and compares results on two benchmarks.
Engineering ·
Wireshark: decrypting HTTPS traffic with TLS 1.3
Published on August 29, 2026, the article explains using SSLKEYLOGFILE and Wireshark to inspect HTTP requests in captured HTTPS traffic.
Engineering ·
Colibri streams MoE experts from disk in pure C
The Colibri project says it can run seven families of open MoE models, from 744 billion to 2.8 trillion parameters, by streaming experts from disk on user hardware.
Engineering ·
Phonon-1 launches for speech recognition
A post dated August 28, 2026, describes Phonon-1, a speech recognition model with 782 million parameters, a 415 MB download, and an Apache 2.0 license.
Engineering ·
Reported 1-bit quantization results for GLM-5.3
In a post dated August 28, 2026, an engineer reports dynamically quantizing GLM-5.3 to 1 bit: 217 GB versus 1.5 TB in BF16, with about 76% top-1 accuracy.
Engineering ·
TurnBench measures turn-taking in spoken dialogue
Published on August 28, 2026, TurnBench includes 30 hours of dyadic speech annotated by three people, a scoring protocol based on conversation analysis, and results for 14 systems.
Engineering ·
WikiSkill reuses knowledge across agent iterations
The approach separates execution traces, a persistent knowledge wiki, and executable skills. Experiences consolidated in the wiki guide later skill updates.
Engineering ·
Sampling to optimize objectives without RL
Published on August 27, 2026, the post opens a series on alternatives to reinforcement learning for optimizing evaluable objectives. Part I introduces sampling methods.
Engineering ·
SlimServe fork enables PCIe P2P on GeForce
A post dated August 27, 2026 reports a graphics-driver change enabling peer-to-peer transfers over PCIe on GeForce cards while optimizing SlimServe. The author has not tested the proposed use on the RTX Pro 6000.
Engineering ·
JIT-Agent adapts agent harnesses to tasks
An article published August 27, 2026 describes JIT-Agent, which synthesizes agent harnesses on demand and can repair and evolve them.
Engineering ·
FireRedAudio combines speech tasks in one model
A record published on August 27, 2026 describes an audio model combining speech recognition, audio understanding, speech synthesis, and editing.
Engineering ·
Architecture overview of GLM-5.3-Flash
A publication dated August 26, 2026 identifies Ox Alpha as GLM-5.3-Flash and outlines components of its hybrid architecture.
Engineering ·
Post describes hybrid attention in GLM-5.3
Published on August 26, 2026, the post claims GLM-5.3 alternates linear and sparse attention, and compares this design with other hybrid approaches.
Engineering ·
Breeze TTS 2 releases open weights
A post dated August 26, 2026, says Breeze TTS 2 weights are open and claims the model leads a ranking of TTS provider voices.
Engineering ·
Practical resources on GRPO and OPD
A publication dated August 26, 2026 shares two resources on post-training with reinforcement learning beyond SFT: one on scaling vanilla GRPO and another on OPD.
Engineering ·
Pocket TTS runs on Android with LiteRT
A publication dated August 26, 2026, reports that the 100-million-parameter Pocket TTS model, with six included voices, runs on-device with LiteRT on current Android devices.
Engineering ·
Recuris proposes agent memory without changing the model
In a publication dated August 26, 2026, Recuris describes two loops for long-horizon agents: working and experiential memories guide tasks, while a fixed meta-agent analyzes execution traces and proposes targeted repairs.