Multiple-Choice Question Answering

Large language models often generate confident but incorrect answers rather than abstaining when uncertain. This problem is particularly acute for small language models (SLMs), where computational constraints and autonomous operation amplify the need for reliable uncertainty detection. We propose Second Guess, a lightweight, parameter-free prompting technique for abstention in multiple-choice question answering (MCQA) that is well-suited for SLMs. Our key empirical insight is that models which truly know an answer will select it consistently, while uncertain models exhibit unstable behavior when an ``I don't know'' option is added. Evaluated on four open models (2B-8B parameters) and four benchmarks, Second Guess achieves the highest composite risk improvement of 10.81%. Notably, it maintains an 8% composite risk improvement on fine-tuned models where entropy-based methods degrade, and improves most for lower-performing models. All code and results required to reproduce this work is available in https://github.com/Mystic-Slice/second-guess

Disagreement between language models is often read as missing knowledge. Questions that need the same missing knowledge should then be right or wrong together. We test for this structure with the eigenvalues of the questions' correlation matrix. Within one model, independently drawn answers show no detectable structure; it appears when all questions in one pass over the benchmark share a random seed or few-shot prompt. Across 24 models, the largest structure beyond overall skill is a preference among answer letters: when the options are rotated, it follows the letter, not the question, and it depends on one model, Llama-3.2-1B; without it only a smaller excess remains. Counting a question correct only if right in every rotation removes the part that acts on guesses. What remains is a marginal excess at the 5% line, tied post hoc to two near-duplicate model pairs and to Llama-3.2-1B.

Multiple-choice QA benchmarks usually evaluate small language models (SLMs) as direct answerers, but deployed language-model systems increasingly rely on external scaffolds such as tools, code, and repeated model calls. We introduce Code-Guided Reasoning (CGR), an evaluation protocol and generated-program resource for measuring when executable reasoning scaffolds improve SLM performance on MCQA tasks. CGR standardizes six components: a normalized item interface, a direct solver prompt, a generator prompt, a Python scaffold, solver-call and extraction helpers, and a three-channel result record. On 20,498 retained result rows from a locally prepared MCQA bundle and six metadata-registered solver models, the observed non-zero-baseline partition shows 66.21% macro assisted accuracy versus 38.11% direct accuracy, a +28.10 percentage-point difference with a pair-bootstrap interval of [20.32, 36.43]. Under a stricter Ab > 30% direct-signal gate, the macro difference is +14.11 points. These estimates are descriptive. Assisted inference uses a larger solver-call budget, answer extraction is brittle, Time-MQA contains the observed regressions, and some generated programs violate the no-hard-coding instruction. CGR provides the trace package needed to interpret these results, including direct, assisted, and generator-side answers, partition definitions, generated programs, response metadata, and audits.

Probing the capabilities of Large Language Models (LLMs) and building robust solutions for Multiple-Choice Question Answering (MCQA) remain central challenges in natural language understanding. Furthermore, the rapid proliferation of LLMs has created the implicit assumption that more sophisticated prompting techniques yield better performance. Several studies claim such gains, but report them under differing models, prompt wordings and answer-extraction rules, so the gains cannot be attributed to the technique alone. We address this gap with ReasonLab, an evaluation framework in which the prompting technique is a first-class experimental variable alongside the model and the dataset, and which retains every generation for inspection. Using ReasonLab we conduct a controlled study of 8 prompting techniques across 10 MCQA datasets, 27 model configurations and 480,927 evaluations at temperature 0. We find that the prompting technique is a minor determinant of accuracy: on configurations without a reasoning budget the reasoning triggers improve on direct prompting by only 3.92 to 4.69 pp and are indistinguishable from one another, and on configurations with reasoning enabled no technique differs by more than 0.51 pp. Self-Generate is the only technique with a consistent effect, a reduction of 2.95 pp. We further investigate three phenomena: (1) the comparison of models on a common set of datasets, where model size does not predict accuracy, (2) the trade-offs across thinking budgets, where enabling reasoning is worth up to 12.74 pp whereas an eightfold budget increase adds only 0.48 to 2.10 pp, and (3) the variation in dataset difficulty, with 60% of benchmarks below 70% accuracy and a 43.9 pp spread from easiest to hardest. These results suggest that, for MCQA, the prompting technique is a minor lever compared with enabling model reasoning, and that substantial headroom remains.

We present a comprehensive evaluation of Large Language Models (LLMs) on Computer Science (CS) Data Structure examination questions. Our work introduces a new benchmark dataset comprising exam questions from Tel Aviv University (TAU), curated to assess LLMs' abilities in handling closed and multiple-choice questions. We evaluated the performance of OpenAI's GPT 4o and Anthropic's Claude 3.5, popular LLMs, alongside two smaller LLMs, Mathstral 7B and LLaMA 3 8B, across the TAU exams benchmark. Our findings provide insight into the current capabilities of LLMs in CS education.

Chain-of-thought (CoT) reasoning and reasoning-tuned models such as DeepSeek-R1 are commonly assumed to reduce shallow heuristic biases by thinking carefully. We test this on position bias in multiple-choice QA and find a different story: within any reasoning-capable model, per-question position bias scales with the length of the reasoning trajectory. Across thirteen reasoning-mode configurations (two R1-distilled 7-8B models, two base models prompted with CoT, and DeepSeek-R1 at 671B) on MMLU, ARC-Challenge, and GPQA, twelve show a positive partial correlation between trajectory length and Position Bias Score (PBS) after controlling for accuracy, ranging from 0.11 to 0.41 (all p < 0.05). All twelve open-weight reasoning-mode configurations show monotonically increasing PBS across length quartiles. A truncation intervention provides causal evidence: continuations resumed from later points in the trajectory are increasingly likely to shift toward position-preferred options (16% to 32% for R1-Qwen-7B across absolute-position buckets). At 671B, aggregate PBS collapses to 0.019, but the length effect still manifests in the longest quartile (PBS = 0.071), suggesting that accuracy gates the expression of length-driven bias rather than eliminating the underlying mechanism. We additionally find that direct-answer position bias is a distinct phenomenon with a different footprint (strong in Llama-Instruct-direct, weak in Qwen-Instruct-direct, and uncorrelated with trajectory length): CoT reasoning replaces this baseline bias with length-accumulated bias. Our results argue that reasoning-capable models should not be treated as order-robust by default in MCQ evaluation pipelines, and offer a diagnostic toolkit (PBS, commitment change point, effective switching, truncation probes) for auditing position bias in reasoning models.

Multiple choice evaluation is widely used for benchmarking large language models, yet near ceiling accuracy in low option settings can be sustained by shortcut strategies that obscure true competence. Therefore, we propose a massive option evaluation protocol that scales the candidate set to one hundred options and sharply reduces the impact of chance performance. We apply this framework to a Korean orthography error detection task where models must pick the single incorrect sentence from a large candidate set. With fixed targets and repeated resampling and shuffling, we obtain stable estimates while separating content driven failures from positional artifacts. Across experiments, results indicate that strong performance in low option settings can overstate model competence. This apparent advantage often weakens under dense interference at high $N$, revealing gaps that conventional benchmarks tend to obscure. We identify two failure modes, semantic confusion and position bias toward early options under uncertainty. To isolate the effect of context length, we run padding controlled and length matched tests, which suggest that the main bottleneck is candidate ranking rather than context length. Together, these findings support massive option evaluation as a general framework for stress testing model reliability under extreme distractor density, beyond what low option benchmarks can reveal.

Retrieval-Augmented Generation (RAG) improves Large Language Models (LLMs) by grounding generation in external, non-parametric knowledge. However, when a task requires choosing among competing options, simply grounding generation in broadly relevant context is often insufficient to drive the final decision. Existing RAG methods typically rely on a single initial query, which often favors topical relevance over decision-relevant evidence, and therefore retrieves background information that can fail to discriminate among answer options. To address this issue, here we propose Hypothesis-Conditioned Query Rewriting (HCQR), a training-free pre-retrieval framework that reorients RAG from topic-oriented retrieval to evidence-oriented retrieval. HCQR first derives a lightweight working hypothesis from the input question and candidate options, and then rewrites retrieval into three targeted queries that seek evidence to: (1) support the hypothesis, (2) distinguish it from competing alternatives, and (3) verify salient clues in the question. This approach enables context retrieval that is more directly aligned with answer selection, allowing the generator to confirm or overturn the initial hypothesis based on the retrieved evidence. Experiments on MedQA and MMLU-Med show that HCQR consistently outperforms single-query RAG and re-rank/filter baselines, improving average accuracy over Simple RAG by 5.9 and 3.6 points, respectively. Code is available at https://anonymous.4open.science/r/HCQR-1C2E.

Multiple-choice questions (MCQs) are widely used to evaluate large language models (LLMs). However, LLMs remain vulnerable to the presence of plausible distractors. This often diverts attention toward irrelevant choices, resulting in unstable oscillation between correct and incorrect answers. In this paper, we propose Inclusion-of-Thoughts (IoT), a progressive self-filtering strategy that is designed to mitigate this cognitive load (i.e., instability of model preferences under the presence of distractors) and enable the model to focus more effectively on plausible answers. Our method operates to reconstruct the MCQ using only plausible option choices, providing a controlled setting for examining comparative judgements and therefore the stability of the model's internal reasoning under perturbation. By explicitly documenting this filtering process, IoT also enhances the transparency and interpretability of the model's decision-making. Extensive empirical evaluation demonstrates that IoT substantially boosts chain-of-thought performance across a range of arithmetic, commonsense reasoning, and educational benchmarks with minimal computational overhead.

Reinforcement Learning with Verifiable Rewards (RLVR) significantly enhances the reasoning capabilities of Large Language Models. When applied to RLVR, Multiple-Choice Questions (MCQs) offer a scalable source of verifiable data but risk inducing reward hacking, where models shortcut reasoning via random guessing or simple elimination. Current approaches often mitigate this by converting MCQs to open-ended formats, thereby discarding the contrastive signal provided by expert-designed distractors. In this work, we systematically investigate the impact of option design on RLVR. Our analysis highlights two primary insights: (1) Mismatches in option counts between training and testing degrade performance. (2) Strong distractors effectively mitigate random guessing, enabling effective RLVR training even with 2-way questions. Motivated by these findings, we propose Iterative Distractor Curation (IDC), a framework that actively constructs high-quality distractors to block elimination shortcuts and promote deep reasoning. Experiments on various benchmarks demonstrate that our method effectively enhances distractor quality and yields significant gains in RLVR training compared to the original data.

In recent years, corpus-driven AI methods, such as Large Language Models (LLMs), have seen widespread use in education. While on the surface their abilities look promising for tasks ranging from generating assessment materials to simulating student performance, we should be aware of the subtle nuances of their frequentist nature that might be affecting their behaviour. In this work, we focus on the aspect of corpus frequency in the context of creating high-quality Multiple Choice Questions (MCQs), specifically asking: What if corpus prevalence were enough to identify the correct answer to an MCQ? We propose a computational method of assessing corpus prevalence of MCQ options in large text corpora leveraging textual embeddings using both expert- and machine-generated MCQ sets. The key finding, across three large question sets, is that correct answers, independently of the question stem, are significantly more available than incorrect options. Specifically, using Wikipedia as the retrieval corpus, we find that always selecting the most prevalent option leads to scores up to 9.0% above the random-guess baseline. We also find that MCQ distractors generated by LLMs often show similar patterns of prevalence compared to expert-created options, despite the LLMs' frequentist nature and their training on large collections of textual data. Moreover, we find that corpus prevalence does not necessarily correlate with how recognisable terms are to humans. This highlights the need to better understand how corpora are used in AI-driven methods for education, whether applied directly or indirectly via LLMs.

Recent advances in large audio language models (LALMs) have primarily been assessed using a multiple-choice question answering (MCQA) framework. However, subtle changes, such as shifting the order of choices, result in substantially different results. Existing MCQA frameworks do not account for this variability and report a single accuracy number per benchmark or category. We dive into the MCQA evaluation framework and conduct a systematic study spanning three benchmarks (MMAU, MMAR and MMSU) and four models: Audio Flamingo 2, Audio Flamingo 3, Qwen2.5-Omni-7B-Instruct, and Kimi-Audio-7B-Instruct. Our findings indicate that models are sensitive not only to the ordering of choices, but also to the paraphrasing of the question and the choices. Finally, we propose a simpler evaluation protocol and metric that account for subtle variations and provide a more detailed evaluation report of LALMs within the MCQA framework.

Large language models (LLMs) routinely fail to output the correct option in multiple-choice question answering (MCQA) while encoding the answer internally. We expose this latent knowledge via the Query--Key (QK) score, defined for an attention head as the inner product between the last-token query and the key at the end-of-line token following option $i$, evaluated before rotary positional embedding is applied. Its argmax identifies a universal class of select-and-copy heads in middle layers that perform option selection through semantic query--key alignment, mechanistically distinct from induction and copy-suppression heads (Olsson et al., 2022): they are invariant to label symbols, and solve a synthetic task with zero surface overlap---properties no positional-copy account explains and that critically require stripping RoPE. Across 24 models from 1.5B to 72B parameters (LLaMA-2/3/3.1/3.3, Qwen-2.5, Gemma, Phi-3.5, DeepSeek-R1-Distill), a single head's QK-score exceeds the model's own zero-shot accuracy by up to $+27.4$ pp on HellaSwag and $+49.8$ pp on HaluDialogue; causal zero-ablation collapses MCQA accuracy to near-random. To remove any dependence on labeled validation data, we introduce an unsupervised HeadScore that ranks heads from unlabeled inputs and recovers the supervised top-$k$ heads on every tested model. Against four positional-debiasing baselines (e.g., PriDe, Wiegrefe, Wang), QK-score is complementary by construction: debiasing re-weights output logits, whereas QK-score reads the model's selection from a middle-layer head before decoding. We release a one-line drop-in HeadScore script and per-model head indices, making every result one-command reproducible across all 24 models and four benchmarks.