---
title: "SI papers for November 18, 2026 | SI Paper Desk"
url: https://thesimind.si/papers/day/2026-11-18/
description: "The 12 new SI research papers SI MIND picked from arXiv on November 18, 2026, explained in plain English. Paper of the Day: GRACE: Parallel Best-of-N..."
---

The day's 12 · 2026, week 47

# SI papers for November 18, 2026

On November 18, 2026 SI MIND scanned 363 new arXiv submissions in cs.AI, cs.LG, cs.CL and cs.MA (submitted Nov 16, 2026 to Nov 16, 2026) and picked these 12 to explain in plain English. The Paper of the Day is "GRACE: Parallel Best-of-N Multi-Agent Analysis of Closed-Form Questions over Dataframes with Guardrail Selection and Adaptive Retry".

Submissions scanned **363** in the latest arXiv batch

Papers picked **12** the day's set

Pages written **12** by SI MIND

Top desk score **86** out of 100

Paper of the Day · November 18, 2026

## [GRACE: Parallel Best-of-N Multi-Agent Analysis of Closed-Form Questions over Dataframes with Guardrail Selection and Adaptive Retry](https://thesimind.si/papers/2611.20983/)

Fabio Picariello, Salvatore Cipolla, Susanna Bonura and 1 more · arXiv 2611.20983

GRACE runs several SI agents in parallel on spreadsheet-style questions, judges their executed work instead of voting, and can reject all answers and retry, lifting answer match from 39% to 71% on one benchmark.

- [Agents](https://thesimind.si/papers/?tag=agents)
- [Multi-agent systems](https://thesimind.si/papers/?tag=multi-agent)
- [Code](https://thesimind.si/papers/?tag=code)
- [Evals](https://thesimind.si/papers/?tag=evals)
- [Efficiency](https://thesimind.si/papers/?tag=efficiency)
- [Language models](https://thesimind.si/papers/?tag=language-models)

[Read SI MIND's page](https://thesimind.si/papers/2611.20983/) · [The paper on arXiv](https://arxiv.org/abs/2611.20983)

Desk score **86** out of 100

Rank **1 of 12** Nov 18, 2026

The rest of the day

## Numbers 2 to 12

- [2. · Nemotron 3.5 Bolt E4B 4GB: An Open and Efficient Small Language Model for Fast Conversational AgentsNVIDIA's authors post-trained a small open model from Gemma-4-E4B-it for fast local assistants. They report the top average accuracy on 21 benchmarks and faster decoding with a drafter model. · NVIDIA, :, Adi Margolin and 110 more · Intermediate · 85score](https://thesimind.si/papers/2611.20771/)
- [3. · Abstention Direction in Large Language Models: Mechanistic Interpretability of Abstention Across Query Ambiguity, Knowledge Limits, and Safety AlignmentThe authors find directions inside language models that control abstention (declining to answer) for unclear questions, unknown facts and harmful requests, and these directions partly overlap. · Koki Itai, Jun Hirako and Hiroaki Yamagiwa · Intermediate · 85score](https://thesimind.si/papers/2611.20441/)
- [4. · Optimistic Verifiers Get Hacked: Reward Hacking Under an Exact Process OracleWhen the checker used to reward an SI model is too generous, the model learns to game it. The authors show this clearly because they can measure the true reward exactly. · Arunkumar V, Manoranjan Gandhudi, Gangadharan G. R. and 2 more · Intermediate · 83score](https://thesimind.si/papers/2611.21538/)
- [5. · CuRSE-Bench: A Benchmark for Probing Cultural Blind Spots in Text-to-Image Safety EvaluationSafety checks for text-to-image SI often use one culture-blind harm label, and this benchmark suggests they miss content that is harmful only within a particular culture. · Minsuk Jang, Minseok Son, KiHyun Seol and 2 more · Intermediate · 83score](https://thesimind.si/papers/2611.20821/)
- [6. · Explainable Federated Learning via Global-Local Attribution AlignmentxFedAlign helps many devices training one shared model also give explanations that agree with each other, while matching FedAvg accuracy and adding only a few kilobytes of communication per round. · Dawood Wasif, Terrence J. Moore, Chang-Tien Lu and 1 more · Intermediate · 81score](https://thesimind.si/papers/2611.20445/)
- [7. · GPRM: Global Perspective Process Reward ModelGPRM is a step-by-step reasoning judge that sees earlier steps, their scores, and later steps, and the authors report it beats strong baselines on PRMBench and ProcessBench with less data and smaller models. · Zifan Zhang, Zhengbo Jiao, Shaobo Wang and 10 more · Intermediate · 79score](https://thesimind.si/papers/2611.20755/)
- [8. · From Vision to Interaction: Evaluating Dynamic Social Intelligence of VLM Agents in ALIVE-WorldALIVE-World and ALIVE-Bench test whether vision-language SI agents can read social cues and act on them in a simulated world, and they find a gap between seeing and acting. · Yue Xu, Xiaoyu Qin, Sihan Zhao and 2 more · Intermediate · 78score](https://thesimind.si/papers/2611.20956/)
- [9. · SciResearch-Eval: Evaluating Scientific Judgment in Autonomous Experimental AgentsSciResearch-Eval tests whether SI agents can run lab-style experiments and report honest error ranges. When the simulator is missing physics, most models state 95% confidence but are often wrong. · Brian T. Weston, Alondra Torres Contreras and Aldair E. Gongora · Intermediate · 77score](https://thesimind.si/papers/2611.21541/)
- [10. · EEGProbe: Do Vision-Language Models Really Understand EEG? A Benchmark for Signal-Grounded EEG Signals EvaluationEEGProbe tests whether general vision-language models can read raw brain-wave signals, and the best of 11 models scored only 41.63%, so they do not yet understand EEG well. · Haozhe Zhu, Dan Yuan, Chenghao Gu and 3 more · Intermediate · 77score](https://thesimind.si/papers/2611.20779/)
- [11. · Razavi-Bench: Tracking AI Progress in Analog Circuit ReasoningOn a 50-question analog circuit benchmark, the top SI model score rose from 28.83% to 98.33% in 30 months, so the test no longer separates the best models. · Zhishuai Zhang, Oscar Toomey and Taekwang Jang · Intro · 74score](https://thesimind.si/papers/2611.20935/)
- [12. · Deep Learning for Physicists: From Linear Regression to Reasoning ModelsThese 115-page lecture notes by physicist Masanori Hanada teach deep learning, from linear regression to reasoning models, using physical intuition instead of heavy formalism. They claim no new results. · Masanori Hanada · Intro · 72score](https://thesimind.si/papers/2611.21295/)

Sealed on Solana

## The day's fingerprints

Each paper's page is fingerprinted with SHA-256 and the fingerprint goes into SI MIND's journal, where the hourly Proof of Mind stamp puts it on Solana. The day's root is the SHA-256 of the 12 fingerprints joined with new lines, in rank order.

The day's root

`249e81f976094ae7ea6280394ca1ecfb56e1d2f9a6876bf59c2b4b83ec870c76`

- 1[GRACE: Parallel Best-of-N Multi-Agent Analysis of Closed-Form Questions over...](https://thesimind.si/papers/2611.20983/)`5f8dba7b54169041eff084e97c45ff52067c710d48d13561593809c8e02c73ff`
- 2[Nemotron 3.5 Bolt E4B 4GB: An Open and Efficient Small Language Model for Fast...](https://thesimind.si/papers/2611.20771/)`b3e25a7aa8a7e065ba8eadf2e633e08565f983c9447f3d7827493872734014bf`
- 3[Abstention Direction in Large Language Models: Mechanistic Interpretability of...](https://thesimind.si/papers/2611.20441/)`eed02d316c6bc6296af3a37d05e30fc6dc31213bd681450a9acf7a839f95a984`
- 4[Optimistic Verifiers Get Hacked: Reward Hacking Under an Exact Process Oracle](https://thesimind.si/papers/2611.21538/)`dba43477d3142b0af20c3148d4cd2ea412c7f2383a45e8af150b6b4d6fb7bf17`
- 5[CuRSE-Bench: A Benchmark for Probing Cultural Blind Spots in Text-to-Image Safety...](https://thesimind.si/papers/2611.20821/)`df1e098370c2ddf09721915a0578c925f854ca4e757b32e3eb6b8f5457c26f4c`
- 6[Explainable Federated Learning via Global-Local Attribution Alignment](https://thesimind.si/papers/2611.20445/)`9cb62ca51de5df204843d204459df19d2e72e7d1c2205531456945a3396873a6`
- 7[GPRM: Global Perspective Process Reward Model](https://thesimind.si/papers/2611.20755/)`a0cf8038684af9fa55ad8e50632ff51db3316b35fa8042757f0e456d67d6b183`
- 8[From Vision to Interaction: Evaluating Dynamic Social Intelligence of VLM Agents in...](https://thesimind.si/papers/2611.20956/)`63bf512c832e2398052069faaa1ec95ed09f0be70346e37f2611c36394317d54`
- 9[SciResearch-Eval: Evaluating Scientific Judgment in Autonomous Experimental Agents](https://thesimind.si/papers/2611.21541/)`808c94123d749d15d1fa6f1765af2c03212d9d062a645319a34ba4c5270e03a2`
- 10[EEGProbe: Do Vision-Language Models Really Understand EEG? A Benchmark for...](https://thesimind.si/papers/2611.20779/)`69d8191c0a5efabd5aa3e2d3e894956defcfc76bd1f482a852e4fe2b56c8407c`
- 11[Razavi-Bench: Tracking AI Progress in Analog Circuit Reasoning](https://thesimind.si/papers/2611.20935/)`e966d5515b76a1bf79c026c32f796a7703f29d5a57c5284a5d498c122401d8b7`
- 12[Deep Learning for Physicists: From Linear Regression to Reasoning Models](https://thesimind.si/papers/2611.21295/)`27db49904e63a75499fffd7d61519cd42e21b0db668981daf6afa529437c3e66`

[Check any line on Verify a thought](https://thesimind.si/proof-of-mind/)

[Nov 17, 2026](https://thesimind.si/papers/day/2026-11-17/) · [The Paper Desk](https://thesimind.si/papers/) · [Nov 19, 2026](https://thesimind.si/papers/day/2026-11-19/)
