Research Radar is an automated sweep of new work in AI/LLM safety, alignment, and pragmatic mechanistic interpretability. Claude Code cloud routines search OpenReview, the ACL Anthology, TMLR, arXiv (cs.CL / cs.LG / cs.CR / cs.AI), and the alignment forums, then write up what they find: a daily top ten Tuesday through Sunday, and a wider aggregate on Monday mornings.
Entries are ranked by importance to that agenda rather than by recency, and peer-reviewed work outranks incremental preprints. Each one carries a technical summary of method and result rather than a reprinted abstract, plus the venue and its peer-review status. If a day turns up fewer than ten genuinely relevant papers, the report says so instead of padding.
Since the last weekly
Window: August 5–7, 2026 · Sources swept: OpenReview, ACL Anthology, TMLR, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), ICML 2026 Mechanistic Interpretability Workshop · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Window: August 5–6, 2026 · Sources swept: OpenReview, ACL Anthology, TMLR, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), LessWrong/Alignment Forum, lab blogs · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Window: August 2–4, 2026 · Sources swept: OpenReview, ACL Anthology, TMLR, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), LessWrong/Alignment Forum, lab blogs · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Window: July 31 – August 2, 2026 · Sources swept: OpenReview, ACL Anthology, TMLR, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), LessWrong/Alignment Forum, lab blogs · Counts: 0 peer-reviewed · 10 preprints · 0 forum/blog
Window: July 30–August 1, 2026 (latest preprints); peer-reviewed sweep back through May 2026 for work not previously covered · Sources swept: OpenReview, ACL Anthology, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, GitHub · Counts: 1 peer-reviewed · 9 preprints · 0 forum/blog
Window: July 29–31, 2026 (latest preprints); back-swept for peer-reviewed work not previously surfaced · Sources swept: OpenReview, ACL Anthology, ESORICS proceedings, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Window: July 27–29, 2026 (latest preprints); back-swept through April 2026 for peer-reviewed work not previously surfaced · Sources swept: OpenReview, ACL Anthology, NeurIPS/ICLR proceedings, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar · Counts: 2 peer-reviewed · 1 workshop paper · 7 preprints · 0 forum/blog
Window: July 26–28, 2026 (latest preprints); back-swept through April 2026 for field-relevant work not previously surfaced · Sources swept: OpenReview, ACL Anthology, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, GitHub · Counts: 1 peer-reviewed · 9 preprints · 0 forum/blog
Window: July 24–26, 2026 (latest preprints); back-swept through April 2026 for field-relevant work not previously surfaced · Sources swept: OpenReview, ACL Anthology, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, GitHub · Counts: 0 peer-reviewed · 10 preprints · 0 forum/blog
Window: July 23–25, 2026 (latest preprints); back-swept through April 2026 for field-relevant work not previously surfaced · Sources swept: OpenReview, ACL Anthology, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, GitHub · Counts: 1 peer-reviewed · 9 preprints · 0 forum/blog
Window: July 21–24, 2026 (latest preprints); back-swept through May 2026 for field-relevant work not previously surfaced · Sources swept: OpenReview, ACL Anthology, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, LessWrong/Alignment Forum · Counts: 1 peer-reviewed · 9 preprints · 0 forum/blog
Window: July 22–23, 2026 (latest preprints); back-swept through June 2026 for field-relevant work not previously surfaced · Sources swept: OpenReview, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, GitHub · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Window: July 20–22, 2026 (latest preprints); back-swept through May 2026 for field-relevant work not previously surfaced · Counts: 0 peer-reviewed · 10 preprints · 0 forum/blog
Window: July 19–21, 2026 (sweeping back ~2 weeks for field-relevant work) · Sources swept: OpenReview, arXiv (cs.CL/cs.LG/cs.CR/cs.AI), Semantic Scholar, HuggingFace Papers, GitHub · Counts: 2 peer-reviewed · 8 preprints · 0 forum/blog
Weekly editions
Window: July 13–22, 2026 · Fresh sweep covers: ICLR 2026 (OpenReview), ICML 2026 (main + Mechanistic Interpretability Workshop), ACL 2026 (main + Findings), NeurIPS 2025, COLM 2025, NeurIPS 2026 Competition Track, arXiv cs.CL/cs.LG/cs.CR/cs.AI · Counts: 8 peer-reviewed (3 ICLR 2026 · 1 NeurIPS 2025 · 1 ICML 2026 Mech Interp Workshop · 3 ACL 2026 main) · 7 preprints · 0 forum/blog
Window: July 6–12, 2026 · Fresh sweep covers: ICLR 2026, NeurIPS 2025, ICML 2026 (main + Mech Interp Workshop), IEEE S&P 2026, USENIX Security 2026, arXiv cs.CL/cs.LG/cs.CR/cs.AI, Transformer Circuits Thread, LessWrong/Alignment Forum · Counts: 10 peer-reviewed (3 ICLR 2026 · 2 NeurIPS 2025 · 1 ICML 2026 Oral · 1 IEEE S&P 2026 · 1 USENIX Security 2026 · 2 ICML 2026 Workshop spotlights) · ~22 preprints · 1 lab blog
Window: June 30 – July 5, 2026 · Fresh sweep covers: ICLR 2026, ICML 2026 (main + Mech Interp Workshop + DL for Code Workshop), ECCV 2026, ACL 2026, arXiv cs.CL/cs.LG/cs.CR/cs.AI, LessWrong/Alignment Forum, lab blogs · Counts: 13 peer-reviewed (3 ICLR 2026, 4 ICML 2026 main, 3 ICML 2026 workshops, 2 ECCV 2026, 1 ACL 2026) · 24 preprints · 0 forum/blog posts new enough to rank
Backfill
Compiled: 2026-06-30 · Window: posted/accepted Jul 2024 – Jun 2026 (seminal older anchors flagged) · Counts: ~27 items · peer-reviewed: SEDD (ICML'24 best paper), MDLM (NeurIPS'24), DiffuLLaMA (ICLR'25), Block Diffusion (ICLR'25 oral), Fast-dLLM (ICLR'26), DIJA/A2D/DiffuGuard (ICLR'26); rest preprints/industry reports.
Compiled: 2026-06-30 · Window: posted or accepted June 2025 – June 2026 · Coverage: ~110 unique items total — Part I ~52 (topic sweeps, §1–§8) + Part II ~58 (completeness additions, §9–§20, incl. the previously-missed AI Control area). ~40 peer-reviewed (incl. NeurIPS/ICML/ICLR orals & spotlights, ACL/EMNLP outstanding, S&P/USENIX distinguished); rest landmark preprints / major lab releases.