Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Find, access, and cite conference papers and proceedings effectively
.claude/skills/brycewang-stanford-conference-proceedings-guide/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 60% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 65% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 87% | 0% |
| case-17 | ✓→✓ | = Same ✓ | 7% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 66% | 0% |
A skill for finding, accessing, and citing conference papers and proceedings. In many fields -- especially computer science, engineering, and HCI -- conferences are the primary venue for publishing cutting-edge research. This guide covers major proceedings databases, search strategies, and citation practices.
| Database | Coverage | Access | |----------|----------|--------| | ACM Digital Library | ACM conferences (CHI, SIGCOMM, KDD, etc.) | Institutional or ACM membership | | IEEE Xplore | IEEE conferences (CVPR, ICRA, INFOCOM, etc.) | Institutional or IEEE membership | | DBLP | CS bibliography, links to proceedings | Free metadata | | Springer LNCS | Lecture Notes in Computer Science series | Institutional | | AAAI Digital Library | AAAI, IJCAI proceedings | Free for AAAI papers | | NeurIPS / ICML / ICLR | ML conference proceedings | OpenReview (free) | | arXiv | Preprints including conference submissions | Free |
Strategy 1 - Search by conference name:
Google Scholar: source:"NeurIPS" "transformer" "attention"
DBLP: Browse conference page -> search within proceedings
Strategy 2 - Search by topic across all venues:
OpenAlex: Filter by type "proceedings-article"
Google Scholar: Use keywords, then check venue names in results
Strategy 3 - Track specific conferences:
Bookmark the conference proceedings page (e.g., openreview.net)
Subscribe to DBLP RSS feeds for specific conference series
Follow conference Twitter/social media accounts for announcementspythondef assess_conference_quality(conference_name: str) -> dict: """ Framework for evaluating conference quality and reputation. Args: conference_name: Name or acronym of the conference """ indicators = { "acceptance_rate": { "top_tier": "< 25%", "mid_tier": "25-40%", "lower_tier": "> 40%", "note": "Check conference website for historical rates" }, "ranking_sources": [ "CORE Conference Ranking (core.edu.au)", "CSRankings.org (CS-specific, based on publication counts)", "Google Scholar Metrics (h5-index for venues)", "CCF Ranking (Chinese Computer Federation, A/B/C tiers)" ], "quality_signals": [ "Program committee reputation and size", "Keynote speaker caliber", "Longevity and consistency of the conference series", "Whether proceedings are indexed in Scopus/WoS", "Industry participation and sponsorship" ] } return indicators
Tier A* (Top): ICML, NeurIPS, CVPR, ACL, SIGCOMM, OSDI, CHI, KDD
Tier A: AAAI, ECCV, EMNLP, ICSE, WWW, CIKM, ICDM
Tier B: COLING, WACV, ICSME, PAKDD, AISTATSTiers vary by subfield. Always check the ranking relevant to your specific area.
1. Author homepages: Many researchers post preprints/camera-ready PDFs
2. arXiv: Conference-accepted papers are often on arXiv
3. OpenReview: NeurIPS, ICLR, and others host papers with reviews
4. Institutional repository: Check the authors' university repository
5. Conference website: Some conferences offer free proceedings
6. OpenAlex: Aggregates metadata and OA links from multiple sourcesWorkshop papers are shorter (4-8 pages), less rigorously reviewed, and represent more preliminary work. They are still citable but carry less weight. When citing, always distinguish:
# Main conference paper:
Author et al. "Title." In Proceedings of NeurIPS 2024.
# Workshop paper:
Author et al. "Title." In Workshop on X at NeurIPS 2024.bibtex@inproceedings{vaswani2017attention, title = {Attention Is All You Need}, author = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, Lukasz and Polosukhin, Illia}, booktitle = {Advances in Neural Information Processing Systems}, volume = {30}, year = {2017} }
@article instead of @inproceedingsAlways use the official proceedings citation provided by the conference or digital library.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-17 | pass→pass | 20,948 | 15,591 | -26% | 1 | 1 | 0% | 3,100 | 3,304 | +7% | 0 | 0 | — |
case-01 | fail→fail | 19,671 | 22,051 | +12% | 1 | 1 | 0% | 2,825 | 4,024 | +42% | 0 | 0 | — |
case-02 | pass→pass | 8,205 | 8,388 | +2% | 1 | 1 | 0% | 1,508 | 2,504 | +66% | 0 | 0 | — |
case-03 | pass→pass | 10,746 | 9,897 | -8% | 1 | 1 | 0% | 1,564 | 2,753 | +76% | 0 | 0 | — |
case-04 | pass→pass | 14,138 | 9,891 | -30% | 1 | 1 | 0% | 2,174 | 2,594 | +19% | 0 | 0 | — |
case-05 | pass→pass | 12,963 | 9,345 | -28% | 1 | 1 | 0% | 1,918 | 2,599 | +36% | 0 | 0 | — |
case-06 | pass→pass | 11,771 | 5,425 | -54% | 1 | 1 | 0% | 1,870 | 2,018 | +8% | 0 | 0 | — |
case-07 | pass→pass | 12,160 | 10,269 | -16% | 1 | 1 | 0% | 1,854 | 2,625 | +42% | 0 | 0 | — |
case-08 | fail→fail | 13,884 | 10,647 | -23% | 1 | 1 | 0% | 2,159 | 2,863 | +33% | 0 | 0 | — |
case-09 | fail→pass | 9,996 | 8,009 | -20% | 1 | 1 | 0% | 1,443 | 2,313 | +60% | 0 | 0 | — |
case-10 | pass→pass | 14,881 | 9,371 | -37% | 1 | 1 | 0% | 2,209 | 2,719 | +23% | 0 | 0 | — |
case-11 | fail→pass | 17,187 | 20,271 | +18% | 1 | 1 | 0% | 2,521 | 4,165 | +65% | 0 | 0 | — |
case-12 | pass→pass | 5,702 | 5,583 | -2% | 1 | 1 | 0% | 913 | 2,043 | +124% | 0 | 0 | — |
case-13 | pass→pass | 11,580 | 11,474 | -1% | 1 | 1 | 0% | 1,763 | 2,637 | +50% | 0 | 0 | — |
case-14 | pass→pass | 10,909 | 10,333 | -5% | 1 | 1 | 0% | 1,709 | 2,608 | +53% | 0 | 0 | — |
case-15 | pass→pass | 17,744 | 16,757 | -6% | 1 | 1 | 0% | 2,721 | 3,756 | +38% | 0 | 0 | — |
case-16 | pass→pass | 3,480 | 3,735 | +7% | 1 | 1 | 0% | 573 | 1,726 | +201% | 0 | 0 | — |
case-18 | fail→pass | 8,856 | 10,883 | +23% | 1 | 1 | 0% | 1,479 | 2,765 | +87% | 0 | 0 | — |
case-19 | pass→pass | 12,268 | 8,897 | -27% | 1 | 1 | 0% | 1,878 | 2,517 | +34% | 0 | 0 | — |
case-20 | pass→pass | 16,175 | 14,140 | -13% | 1 | 1 | 0% | 2,788 | 3,646 | +31% | 0 | 0 | — |
case-21 | pass→pass | 17,262 | 19,076 | +11% | 1 | 1 | 0% | 2,821 | 4,233 | +50% | 0 | 0 | — |
case-22 | pass→pass | 18,034 | 20,555 | +14% | 1 | 1 | 0% | 3,089 | 4,645 | +50% | 0 | 0 | — |
case-23 | pass→pass | 5,681 | 4,852 | -15% | 1 | 1 | 0% | 903 | 2,030 | +125% | 0 | 0 | — |
case-24 | pass→pass | 8,374 | 8,277 | -1% | 1 | 1 | 0% | 1,039 | 2,350 | +126% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 24 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.