Agent Capabilities and Harness Engineering — Why Agents Are Systems, Not Just Models Highlights: In Graham Neubig’s segment of CMU 11-768: AI Agents (taught with Daniel Fried), the argument is that assembling an agent is easy while making one work reliably is not. We walk through the capabilities that separate the two — tool calling, long-context coherence, customizability, task management, environment understanding, and safety — and the two levers available for each: harness engineering and…
Read more

What Are AI Agents and How Do They Work? — From Tool Calls to the ReAct Loop Highlights: The opening lecture of CMU’s 11-768: AI Agents (Fall 2026), taught by Daniel Fried and Graham Neubig, sets out what an agent actually is and where current systems break down. We walk through the anatomy of an agent — environment, observations, actions, reward — and follow the path from plain next-token prediction to tool calls, chat templates,…
Read more

Fame and Uniqueness: Measuring What Makes a Cover Memorable Highlights: Romaniuk’s Distinctive Brand Asset framework (2018) splits brand memorability into two axes — Fame (exposure history) and Uniqueness (visual distinctiveness) — but only one of them is recoverable from a picture. We build a full measurement pipeline on top of Gemini 2.5 Pro: genuine pairwise visual-distinctiveness judgments aggregated with the Bradley–Terry model (not a single holistic LLM call), a 13-family literature-grounded absolute-scoring taxonomy, and real…
Read more

Visual Complexity Without an LLM Judge — Head-to-Head Against Gemini Highlights: LLM/VLM pairwise judges have a structural position bias — the first-shown candidate wins more often than it should, worst-case exactly on close calls. We replace the LLM judge with a transparent, CPU-only alternative: handcrafted complexity features + logistic regression on the signed difference \( \mathbf{d} = f(A) – f(B) \). The same fit gives you a per-image score \( \text{score}(x) = \mathbf{w} \cdot f(x)…
Read more

Highlights: In this post we try to answer one practical question: can we score the visual complexity of an image without humans in the loop? We grab the SAVOIAS dataset (1,420 images with human-rated complexity), focus on the 200 advertisements, and try five different recipes — zero-shot CLIP prompts, Ridge regression on CLIP embeddings, pairwise ranking on embedding differences, kNN retrieval on cosine similarity, and finally Gemini 2.5 Pro as a chain-of-thought judge. The headline:…
Read more

Highlights: An image is just a matrix — but the Transformer eats sequences of vectors. The whole “vision” trick of ViT lives in how that matrix is turned into a sequence. We cut a 224×224 image into a fixed 14×14 grid of 16×16 patches, flatten each, and project it through ONE learned Linear layer. Patches are not sampled — every cell of the grid becomes a token, in fixed order. We prepend a learnable [CLS]…
Read more

Highlights: Jason Weston traces the arc from early neural language models to self-improving LLMs that generate their own training data and evaluate their own reasoning System 1 vs System 2: fixed-compute pattern-matching vs deliberate multi-step reasoning — and why the same LLM implements both Chain-of-Thought prompting: adding “Let’s think step by step” jumps GSM8K accuracy from ~10% to 40–50%; few-shot CoT hits 90%+ on MultiArith CoVe + S2A: Chain-of-Verification reduces hallucinations 3× on knowledge list…
Read more

Highlights: Yu Su’s guest lecture in the UC Berkeley CS294-280 course argues language agents are not “LLM + tools” but a new evolutionary stage of machine intelligence. We walk through the agent-first framing and three concrete research pillars — long-term memory (HippoRAG), implicit reasoning (Grokked Transformers), and model-based planning (WebDreamer) — that map directly to classical AI problems re-examined through the lens of LLMs. Agent-first framing: token generation is itself an action; the inner monologue…
Read more

Highlights: Can we measure spatial composition in living room photographs? We score 100 interior images using saliency-based rule-of-thirds alignment, Gemini Vision layout ratings, and CLIP composition prompts — then cross-correlate with color scores from #018 to find rooms that nail both color and layout. Method 1 — Rule of Thirds + Balance: gradient saliency → Gaussian-weighted power point alignment + visual balance index Method 2 — Gemini Vision Layout: send each image to Gemini 2.5…
Read more

Highlights: Can three completely independent methods agree on which living room has the best colors? We rank 500 interior images using Cohen-Or harmonic templates, Hasler-Süsstrunk colorfulness, and CLIP IQA with 44 color-focused prompts — then measure whether they correlate at all. Method 1 — Cohen-Or: K-means palette → saturation-weighted hue histogram → sweep 7 harmonic templates × 36 rotations → H/T/S composite Method 2 — Hasler-Süsstrunk: opponent channels (rg, yb) → colorfulness + 4×4 spatial…
Read more

Highlights: Can you score color quality algorithmically? Not as taste — as math. This post builds a scoring system from first principles: two adjacent color squares, then triplets, then a real room with three spatial regions. We walk through every formula with brand and flag examples you already know, then prove that geometry alone can move the score by five points on an identical palette. Four pair scoring dimensions: contrast (WCAG luminance), harmony (hue peaks),…
Read more

Highlights: Every computer vision pipeline that touches color starts with the same mistake: using RGB. RGB is built for screens, not for human perception. In this post we build a complete color scoring system for living room images — from the right color space (Lab), through palette extraction (K-means), to a two-color harmony scorer tested on 10 global brand palettes. We discover why luxury brands deliberately score low, and what that means for your model.…
Read more