📻 Radio Frequency - 2601

🐍Mamba SSM Academic

arxiv.org·

I Thought LoRA Was Just Cheap Fine-Tuning. This Paper Proved Me Wrong

🧩Cognitive Science

pub.towardsai.net

Auditing Training Data in Domain-adapted LLMs: LoRA-MINT

📝Computational Linguistics Academic

arxiv.org·

USB Devices

🛡️Penetration Testing Reference

docs.orbstack.dev··Hacker News

msalsas/amanuensis: A local-first AI persona for Mastodon and Bluesky — drafts by machine, approved by a human, nothing it can't ground gets published.

🌙Lua Code

github.com··Hacker News

The Fine-Tuning Trap: Evaluating Negative Transfer and the Role of PEFT in Sub-1B Mathematical Reasoning

🌱Minimal ML Academic

arxiv.org·

Measuring Embedding Drift: Why Hybrid Search Saves Stale Models.

💬Prompt Engineering

pub.towardsai.net

the sky, on your ceiling

🔍Lens Libraries

skylightceiling.com·

ArturR0k3r/AkiraOS: AkiraOS is a modular, security-focused embedded platform for resource-constrained devices. Built on Zephyr RTOS with WebAssembly (WASM) runtime support and OCRE container technology

🔌Microcontrollers Code

github.com··Hacker News

AuRA: Internalizing Audio Understanding into LLMs as LoRA

🐍Mamba SSM Academic

arxiv.org·

How Small Can You Go? LoRA Fine-Tuning 270M-8B Models for Merchant Information Extraction in Financial Transactions

💬Prompt Engineering Academic

arxiv.org·

The Philosophy of the Out-of-Office Email

💬Smalltalk VMs News

theatlantic.com

··Hacker News

Less-relevant results

Apple WWDC On-Device AI Deep Dive - Google Docs

🤖AI

gist.is··Hacker News

Subtitle-Aligned Fine-Tuning of Whisper for Swiss German ASR: Benchmark Contamination, Convention Mismatch, and an Honest Baseline at 25.6% WER (13.8% cWER)

🔍Tokenizers Academic

arxiv.org·

Camera for Radio Waves using 4x4 MIMO

brunokeymolen/lora: LoRa (Long Range) communication related projects

MoonRF Updates

Instruction Finetuning DeepSeek-R1-8B Model Using LoRA and NEFTune

Sea Five Uses Dual ESP32-C5 Modules for GPS-Enabled Wardriving

Fine-tune FLUX.2 [Klein] with a LoRA under 60 minutes

Recoverable but Not Stationary:Local Linear Structures in Weights and Activations

I Thought LoRA Was Just Cheap Fine-Tuning. This Paper Proved Me Wrong

Auditing Training Data in Domain-adapted LLMs: LoRA-MINT

USB Devices

msalsas/amanuensis: A local-first AI persona for Mastodon and Bluesky — drafts by machine, approved by a human, nothing it can't ground gets published.

The Fine-Tuning Trap: Evaluating Negative Transfer and the Role of PEFT in Sub-1B Mathematical Reasoning

Measuring Embedding Drift: Why Hybrid Search Saves Stale Models.

the sky, on your ceiling

ArturR0k3r/AkiraOS: AkiraOS is a modular, security-focused embedded platform for resource-constrained devices. Built on Zephyr RTOS with WebAssembly (WASM) runtime support and OCRE container technology

AuRA: Internalizing Audio Understanding into LLMs as LoRA

How Small Can You Go? LoRA Fine-Tuning 270M-8B Models for Merchant Information Extraction in Financial Transactions

The Philosophy of the Out-of-Office Email

Apple WWDC On-Device AI Deep Dive - Google Docs

Subtitle-Aligned Fine-Tuning of Whisper for Swiss German ASR: Benchmark Contamination, Convention Mismatch, and an Honest Baseline at 25.6% WER (13.8% cWER)