Add robot-arm and computer-use demos (videos, results) to the model card

Model card: Watch it think (Minesweeper, Connect Four, Wordle, Sudoku videos); one-move puzzles and games with and without thinking

System 1 only by default (thinking: "off"); adaptive thinking is opt-in

serve_decide.py: answer read-out that works with speculative decoding (+ fallback), system2_only; serve.sh: --max-logprobs 256, MTP=1 option

Model card: thinking on Knowledge & Reasoning only; sample of five other-area benchmarks with adaptive thinking

Model card: Decision Index table first

Model card: Decision Index at the top

Decision Index 0.2.1 with adaptive thinking on all ten Knowledge & Reasoning benchmarks: 62.48 (MMLU-Pro 84.6, BBH 92.0); latency with and without speculative decoding

Model card: chance baselines and DI skill per Knowledge & Reasoning benchmark; HLE stays at chance; ChessBench truncation

Decision Index 0.2.1 with adaptive thinking (61.15; 8 of 10 Knowledge & Reasoning benchmarks adaptive), latency, reasoning controls (thinking default auto, think_budget, chat_template_kwargs)

Model card: remove the System 1 Decision Index score until it is recomputed with adaptive thinking

Add files using upload-large-folder tool

initial commit