AI Explained
Plain explanations of trending AI concepts, with live visualizations.
LoopCoder-v2: two loops of a shared block beat deeper looping — Weight-tied block looping — What does it mean?
Running one shared transformer block twice lifts a 7B model from 43.0 to 64.4 on SWE-bench — but three or more loops regress, a non-monotonic sweet spot.
ConSA learns where to put full vs sliding-window attention per head — Controllable attention sparsity — What does it mean?
ConSA learns, under a fixed budget, which attention heads get full attention and which get a cheap sliding window — instead of hand-coded rules.
AnchorKV makes KV-cache compression safety-aware — Refusal anchor in the KV cache — What does it mean?
AnchorKV gives KV-cache eviction a safety-aware penalty, so compressing the cache to save memory no longer quietly erodes a model's alignment.
AMD ATOM + ATOMesh — Prefill/decode disaggregation on ROCm — What does it mean?
AMD's new ROCm serving stack splits LLM inference into a compute-bound prefill phase and a memory-bound decode phase, running each on its own pool of GPUs.
Variable-width transformers cut FLOPs 22% — Hourglass layer width — What does it mean?
A transformer with wide outer layers and a narrow middle — an hourglass — cuts 22% of FLOPs and 15% of KV-cache at matched loss.
Ternary Mamba compresses an SSM 3.6x with 4 GPU-hours of QAT — Ternary quantization-aware training — What does it mean?
Ternary Mamba forces every weight to -1, 0, or +1 and trains on that grid, shrinking a Mamba SSM 3.6x in 4 GPU-hours.
SoftMoE replaces top-k expert routing — Differentiable soft top-k routing — What does it mean?
SoftMoE swaps a mixture-of-experts' hard top-k router for a differentiable soft top-k, so the router learns straight from the task loss.
SubQ 1.1 Small hits a 12M-token context — Subquadratic sparse attention — What does it mean?
SubQ 1.1 trades dense n² attention for a near-linear sparse form — a 12M-token context, 56× faster than FlashAttention-2 at 1M.
VibeThinker-3B hits 94.3 on AIME 2026 — Diversity-driven RL — What does it mean?
Diversity-driven RL keeps a model's solution strategies wide instead of collapsing onto a few — how a 3B model reaches 94.3 on AIME 2026.
AdaSR teaches LLMs to reason mid-stream — Streaming reasoning — What does it mean?
AdaSR trains LLMs to reason while input still streams in, then deliberate once it lands — split-phase RL credit under a latency-aware reward.
VIA-SD speeds up speculative decoding 10–20% — Tiered confidence-gated verification — What does it mean?
VIA-SD adds a confidence-gated middle tier to speculative decoding — close calls go to a slim sub-network of the same model, not a full re-run.
MaxProof clears IMO/USAMO gold — Defense-in-depth generative verifier — What does it mean?
MaxProof tunes its proof verifier for a very low false-positive rate, so sampling many candidate proofs and picking a winner by tournament actually works.
