# AdvisorQA: Towards Helpful and Harmless Advice-seeking Question Answering with Collective Intelligence

arXiv: https://arxiv.org/abs/2404.11826 · HTML: https://mlsanity.com/paper/2404.11826/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 42 of its 82 references that are in the map.
- First posted on arXiv 18 Apr 2024.

## Builds on (most cited first, 30 of 42)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [Scaling Instruction-Finetuned Language Models](https://mlsanity.com/paper/2210.11416/) (2022)
- [A Call for Clarity in Reporting BLEU Scores](https://mlsanity.com/paper/1804.08771/) (2018)
- [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://mlsanity.com/paper/2304.01373/) (2023)
- [BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models](https://mlsanity.com/paper/2104.08663/) (2021)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [Aligning AI With Shared Human Values](https://mlsanity.com/paper/2008.02275/) (2020)
- [BARTScore: Evaluating Generated Text as Text Generation](https://mlsanity.com/paper/2106.11520/) (2021)
- [P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks](https://mlsanity.com/paper/2110.07602/) (2021)
- [ELI5: Long Form Question Answering](https://mlsanity.com/paper/1907.09190/) (2019)
- [RewardBench: Evaluating Reward Models for Language Modeling](https://mlsanity.com/paper/2403.13787/) (2024)
- [Texygen: A Benchmarking Platform for Text Generation Models](https://mlsanity.com/paper/1802.01886/) (2018)
- [DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts](https://mlsanity.com/paper/2105.03023/) (2021)
- [Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Merging](https://mlsanity.com/paper/2310.11564/) (2023)
- [LEVER: Learning to Verify Language-to-Code Generation with Execution](https://mlsanity.com/paper/2302.08468/) (2023)
- [FLASK: Fine-grained Language Model Evaluation based on Alignment Skill Sets](https://mlsanity.com/paper/2307.10928/) (2023)
- [Hurdles to Progress in Long-form Question Answering](https://mlsanity.com/paper/2103.06332/) (2021)
- [A Theoretical Analysis of NDCG Type Ranking Measures](https://mlsanity.com/paper/1304.6480/) (2013)
- [A Critical Evaluation of Evaluations for Long-form Question Answering](https://mlsanity.com/paper/2305.18201/) (2023)
- [When the Majority is Wrong: Modeling Annotator Disagreement for Subjective Tasks](https://mlsanity.com/paper/2305.06626/) (2023)
- [Reward-Augmented Decoding: Efficient Controlled Text Generation With a Unidirectional Reward Model](https://mlsanity.com/paper/2310.09520/) (2023)

## Built on by (most cited first, 12 of 12)

- [ELEPHANT: Measuring and understanding social sycophancy in LLMs](https://mlsanity.com/paper/2505.13995/) (2025)
- [SimpleToM: Exposing the Gap between Explicit ToM Inference and Implicit ToM Application in LLMs](https://mlsanity.com/paper/2410.13648/) (2024)
- [Are LLM-Judges Robust to Expressions of Uncertainty? Investigating the effect of Epistemic Markers on LLM-based Evaluation](https://mlsanity.com/paper/2410.20774/) (2024)
- [CausalArmor: Efficient Indirect Prompt Injection Guardrails via Causal Attribution](https://mlsanity.com/paper/2602.07918/) (2026)
- [Drift: Decoding-time Personalized Alignments with Implicit User Preferences](https://mlsanity.com/paper/2502.14289/) (2025)
- [Guaranteed Generation from Large Language Models](https://mlsanity.com/paper/2410.06716/) (2024)
- [Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence](https://mlsanity.com/paper/2510.01395/) (2025)
- [IP-Dialog: Evaluating Implicit Personalization in Dialogue Systems with Synthetic Data](https://mlsanity.com/paper/2506.02449/) (2025)
- [DialDefer: A Framework for Detecting and Mitigating LLM Dialogic Deference](https://mlsanity.com/paper/2601.10896/) (2026)
- [Pro-AI Bias in Large Language Models](https://mlsanity.com/paper/2601.13749/) (2026)
- [Micro Language Models Enable Instant Responses](https://mlsanity.com/paper/2604.19642/) (2026)
- [BenSyc: Benchmarking Conversational Sycophancy and Human Alignment in LLMs for Bengali Contexts](https://mlsanity.com/paper/2606.10061/) (2026)

## Abstract (arXiv)

As the integration of large language models into daily life is on the rise, there is a clear gap in benchmarks for advising on subjective and personal dilemmas. To address this, we introduce AdvisorQA, the first benchmark developed to assess LLMs' capability in offering advice for deeply personalized concerns, utilizing the LifeProTips subreddit forum. This forum features a dynamic interaction where users post advice-seeking questions, receiving an average of 8.9 advice per query, with 164.2 upvotes from hundreds of users, embodying a collective intelligence framework. Therefore, we've completed a benchmark encompassing daily life questions, diverse corresponding responses, and majority vote ranking to train our helpfulness metric. Baseline experiments validate the efficacy of AdvisorQA through our helpfulness metric, GPT-4, and human evaluation, analyzing phenomena beyond the trade-off between helpfulness and harmlessness. AdvisorQA marks a significant leap in enhancing QA systems for providing personalized, empathetic advice, showcasing LLMs' improved understanding of human subjectivity.