Cover image

LoRA, But Make It Legible: How CARLoS Turns Chaos into Retrieval Signal

LoRA marketplaces have a familiar business problem hiding inside an unfamiliar technical wrapper: the shelf labels are terrible. A creator uploads an adapter with a catchy name, a handful of sample images, maybe a description, maybe not. A user searches for “vibrant colors,” “pencil sketch,” “cyberpunk lighting,” or “kimono inspired.” The platform returns whatever its text search thinks is nearby. Sometimes that works. Often it does the digital equivalent of recommending a “Coloring Book” LoRA when the user wanted a graphite sketch. Charming, in the same way a vending machine full of unlabeled cans is charming. ...

December 10, 2025 · 17 min · Zelina
Cover image

When Models Teach Themselves: Inside the Rise of SuperIntelliAgent

Image generators fail in very ordinary ways. A prompt asks for a green banana and a blue vase. The model gives you something banana-adjacent, vase-adjacent, and chromatically negotiable. A designer asks for a bowl containing a pizza. The model places the pizza beside the bowl, halfway inside the bowl, or in a bowl-like universe where geometry has apparently resigned. A product team then does the usual dance: collect bad outputs, ask users what they preferred, curate examples, fine-tune later, and call the whole thing “continuous improvement” because the spreadsheet had a date column. ...

December 1, 2025 · 16 min · Zelina
Cover image

From Tadpole to Titan: How DEVFT Grows LLMs Like a Brain

TL;DR for operators Federated LLM fine-tuning sounds attractive until someone asks the rude operational question: who is actually paying for the compute, memory, and communication on the devices? The paper behind DevFT proposes a useful answer: do not fine-tune the full model end-to-end from the first round. Start with a compact submodel, train it federatively, transfer the learned LoRA parameters forward, then expand the model in stages until it reaches the full target size.1 The authors call this Developmental Federated Tuning, and yes, the developmental psychology metaphor is a little enthusiastic. Fortunately, the mechanism is more interesting than the metaphor. ...

August 4, 2025 · 16 min · Zelina
Cover image

Merge Without Mayhem: How Orthogonal Deltas Could Revolutionize Model Composition

TL;DR for operators Model composition usually sounds harmless until someone asks the obvious production question: “Can we remove that client-specific update without retraining the whole thing?” At that point, many elegant AI stacks quietly become sedimentary rock. The MDM-OC paper proposes a cleaner model lifecycle: keep a shared base model, express every fine-tuned specialist as a task delta, orthogonalize those deltas so they interfere less, merge them with tuned coefficients, and subtract a selected delta later when a capability, customer, or data source needs to be removed.1 The important claim is not “we found another averaging recipe.” The claim is that model updates can be treated as separable components in parameter space. ...

August 2, 2025 · 20 min · Zelina
Cover image

The LoRA Mirage: Why Lightweight Finetuning Isn't Lightweight on Privacy

TL;DR for operators Adapters look small. The privacy surface is not. The paper behind LoRA-Leak argues that LoRA fine-tuning does not magically protect the records used to specialise a language model.1 Even though LoRA trains only low-rank adapter weights while leaving the base model frozen, the resulting model can still leak membership information: an attacker may infer whether a given sample was part of the fine-tuning dataset. ...

July 25, 2025 · 17 min · Zelina
Cover image

OmniAvatar’s Metrics & Training: Under the Hood of Next-Gen Avatars

TL;DR for operators OmniAvatar is best read as a shift from “make the mouth move” to “make the person perform.” The paper introduces an audio-driven avatar video generation system that takes a reference image, an audio clip, and a text prompt, then generates facial and semi-body video with synchronised speech, adaptive body motion, and prompt-controlled scene elements.1 ...

June 24, 2025 · 16 min · Zelina
Cover image

Blind Trust, Fragile Brains: Why LoRA and Prompts Need a Confidence-Aware Backbone

TL;DR for operators LoRA and prompts are attractive because they make model adaptation feel almost too easy: add a few examples, attach a small adapter, nudge the model into a domain, and call it customised. The uncomfortable part is that adaptation changes not only what a model says, but how confidently it says it. A compliance assistant that becomes slightly more domain-specific but far more overconfident has not been improved. It has been promoted beyond its competence, a classic corporate move. ...

March 25, 2025 · 14 min · Zelina