Near-optimal vector quantization from Google's ICLR 2026 paper — 95% recall, 5x compression, zero preprocessing, pure Python FAISS replacement

(ICLR 2026 Oral 🔥) Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)

First open-source KVTC implementation (NVIDIA, ICLR 2026) -- 8-32x KV cache compression via PCA + adaptive quantization + entropy coding

[ICLR 2026] Tree-Sliced Sobolev IPM

Code for locating "critical neurons" in LLMs. We show that masking as few as 3 neurons can cripple a model's capabilities (ICLR 2026).

[ICLR 2026] Revisiting Tree-Sliced Wasserstein Distance Through the Lens of the Fermat–Weber Problem

🔬 Official implementation of ExPO-HM: Learning to Explain-then-Detect for Hateful Meme Detection (ICLR 2026). Novel multimodal RL approach for interpretable and explainable content moderation.

Near-optimal vector quantization for LLM KV cache compression. Python implementation of TurboQuant (ICLR 2026) — PolarQuant + QJL for 3-bit quantization with minimal accuracy loss and up to 8x memory reduction.

AI agent skill implementing Google's TurboQuant compression algorithm (ICLR 2026) — 6x KV cache memory reduction, 8x speedup, zero accuracy loss. Compatible with Claude Code, Codex CLI, and all Agent Skills-compatible tools.

Interactive Benchmarking Tool for TurboQuant KV Cache Compression. Supports 2-4 bit quantization with Real-time Metrics

Trace and debug AI agent behavior locally with a step-by-step visual tool that stores data offline for clear inspection and faster development.

ICLR 2026 paper on state-dependent reward shaping with multi-view videos and ViCLIP for reinforcement learning.

🌐 Discover semantically similar AI research papers from ICLR 2026 and other top venues, simplifying your search for relevant academic content.

Visualize and compare reinforcement learning algorithms for LLM training with interactive formulas, pipeline flow, and algorithm metrics.