paperlist.ai

AI conference papers

AI conference paper library

Search papers from 30+ top conferences, including NeurIPS, ICLR, ICML, CVPR, ACL and AAAI. Filter by keyword, then open any paper for its abstract, authors and links.

arXiv

From Understanding to Erasing: Towards Complete and Stable Video Object Removal

Dingming Liu, Wenjing Wang, Chen Li, Jing Lyu, Haotian Dong

Video object removal aims to erase target objects while reconstructing visually plausible and temporally coherent content. However, target objects often induce shadows, reflections, illumination changes and other effects that extend beyond the provided mask, making conventional mask-conditioned completion prone to visible residuals. We therefore formulate side-effect-aware object removal as an understanding-guided process that integrates object--effect relations, affected-region localization, and context-aware reconstruction. Specifically, we introduce Object-Induced Relation Distillation to transfer token-level object--effect relations from a pretrained vision foundation model to the video diffusion model. We then design Object-aware Framewise Context Cross-Attention to combine target-object semantics with per-frame background context for removal and reconstruction, and propose Attention-guided Region Localization to derive a soft spatial prior over the target object and its affected regions. Extensive experiments across multiple benchmarks demonstrate that our method achieves more complete object-and-effect removal and outperforms existing approaches in removal quality, side-effect suppression, and temporal consistency.

arXiv

Place-it-R1: Unlocking Environment-aware Reasoning Potential of MLLM for Video Object Insertion

Bohai Gu, Taiyi Wu, Dazhao Du, Jian Liu, Shuai Yang, Xiaotong Zhao, Alan Zhao, Song Guo

Video object insertion is fundamental to video editing, yet existing diffusion methods often produce visually plausible but physically inconsistent results. We present Place-it-R1, an end-to-end framework for physically plausible video object insertion driven by environment-aware MLLM reasoning. Rather than treating reasoning as a generic text prompt, Place-it-R1 uses the MLLM to analyze the target environment, infer object-scene interactions, and determine where an insertion is physically valid. The resulting reasoning is translated into two complementary forms of guidance for video diffusion: semantic guidance that describes the intended physical interaction and spatial guidance that provides a valid insertion region in each frame. To further align generation with local physical realism, we introduce Spatial Direct Preference Optimization, which leverages an MLLM to rank generated candidates, and introduces a region-aware preference objective that explicitly localizes physical-violation penalties to the inserted-object region. Place-it-R1 further offers flexible and standard modes to trade off environment adaptation and scene preservation. Extensive experiments show that Place-it-R1 produces more physically coherent and visually natural insertions than state-of-the-art methods and achieves competitive results against commercial systems.

arXiv

Recovering Lesion Parameters from Aphasic Picture Naming Error Profiles in Large Language Models

Yong Yang, Roger Newman-Norlund, Xiang Guan, Saeed Ahmadi, Regan Willis, Nadra Salman, Kalil Warren, Sophie Arheix-Parras, Srihari Nelakuditi, Leonardo Bonilha, Christopher Rorden, Rutvik H. Desai, Julius Fridriksson

Interpretability methods for large language models (LLMs) describe internal state but do not directly test whether that state is causally sufficient to produce the observed behavior. In earlier work, we lesioned LLMs to produce error profiles in picture naming, a central task for assessing aphasia, and found that specific lesions produced errors resembling those of individual stroke survivors. Here we ask the inverse question: given an error profile, can the lesion parameters that produced it be recovered, and what does this inverse problem reveal about transformer computation? Lesions in LLaVA-Vicuna 13B were parameterized by layer index, modification percentage, and noise sigma across 4,840 configurations, and error profiles were characterized by a seven-category clinical taxonomy (correct, semantic, unrelated, formal, mixed, neologism, no-response). We trained a multi-task neural network to map error profiles back to perturbation parameters. The problem admitted a partial solution: across 10 independently trained inverse models, modification percentage and noise sigma were recoverable, whereas layer index was recoverable only within a neighborhood. In counterfactual validation, a fresh model instance perturbed with the recovered parameters reproduced the target behavior in 81.4% of cases. This dissociation between low layer recovery and high counterfactual fidelity is consistent with functional redundancy across transformer layers, a property not captured by standard interpretability methods. As an out-of-distribution test, we applied the trained model to picture-naming error profiles from 278 stroke survivors; recovered parameters were syndrome-discriminative, most strongly for perturbation intensity, indicating generalization beyond the training distribution. Counterfactual validation provides a general framework for LLM interpretability claims beyond inverse mapping.

arXiv

NTDH: Complex Reasoning for Comprehensive Affective Analysis

Tianlei Zhu, Zhiwei Liu, Yuyan Wang, Xiao-Yang Liu, Sophia Ananiadou

Comprehensive affective analysis is challenging for two reasons: it spans heterogeneous prediction tasks with continuous, ordinal, and multi-label outputs, and affective meaning is context-dependent, requiring conflicting cues to be reconciled rather than mapped directly to labels. Existing methods learn this mapping directly and do not model the reconciliation explicitly. We recast the task as a complex-reasoning problem, which yields one output interface across heterogeneous label spaces and a trajectory over which a verifiable reward can be optimised; to our knowledge, this is the first such treatment covering both sentiment and emotion. The obstacle is on the data side: affective reasoning traces must be synthesised, and generic synthesis is misaligned with the targets, tolerances, and phenomena of affect, and discards or leaks its failure cases. We propose NTDH, which addresses these four failures. Naturalisation sets the training answer to the gold label, so it is correct by construction. A Tolerance-aware gate checks each answer against the task's own scoring margin. Domain-aware strategies refine the reasoning using ideas from affective science. Directional Hints report only the type and direction of an error, without exposing the target. We train Qwen3-8B with SFT and then GRPO under the same tolerance used for verification (up to a more permissive construction gate on the multi-label subtask), and a component ablation quantifies the data-quality effect of each part. Using 16,302 training records, about 14x fewer than comparable instruction-tuned systems, the final policy improves over its SFT checkpoint on five of six official-test metrics and achieves the strongest EI-reg result among the compared systems, at a Pearson correlation of 0.862.

arXiv

Multi Codec Discrete Diffusion Model for Text Guided Speech Inpainting and Editing

Iftach Shoham, Tali Dror, Oren Gal, Haim Permuter, Gilad Katz, Eliya Nachmani

Speech recordings often contain missing, corrupted, or incorrect regions that must be reconstructed or modified without re-synthesizing the entire utterance. Speech inpainting restores missing segments, whereas speech editing replaces spoken content according to an edited transcript. Both tasks require the generated speech to express the intended words while remaining consistent with the surrounding speaker identity, prosody, timing, and recording conditions. Discrete diffusion is particularly well suited to these tasks because it can iteratively refine masked tokens while jointly conditioning on both left and right acoustic context. We introduce SIEDD, a discrete diffusion framework for text-guided speech inpainting and editing over hierarchical codec tokens. Its core architecture, HiCoDD, follows the RVQ generation order by representing previously generated codebooks as clean, committed acoustic context and applying diffusion only to the current refinement codebook. This separation enables leakage-free joint training while matching sequential coarse-to-fine inference. The model further combines phoneme-level conditioning, span-localized classifier-free guidance, and duration prediction to support both fixed-duration inpainting and variable-duration text edits. On the RealEdit benchmark, SIEDD achieves the best overall speech-editing performance among the evaluated methods. It also outperforms the evaluated autoregressive baselines across all speech-inpainting settings, on both single and multiple gaps. These results demonstrate that explicitly modeling the codec hierarchy substantially improves context-preserving speech reconstruction and editing. See our full code at https://github.com/iftachShoham/SIEDD.

arXiv

Neural Diversity Regularizes Hallucinations in Language Models

Kushal Chakrabarti, Nirmal Balachundhar

Language models continue to hallucinate despite increases in parameters, compute, and data. We propose neural diversity -- decorrelated parallel representations -- as a principled mechanism that reduces hallucination rates at fixed parameter and data budgets. While existing mitigation strategies largely target accuracy, we provide the first formal tail bounds for hallucination probability in ensembled language models, reframing it as a second-moment reliability problem and explaining 94.3% of empirical reliability variation seen across parallel configurations. We introduce ND-LoRA (Neural Diversity Low-Rank Adaptation), combining parallel LoRA adapters with Barlow Twins regularization, and reduce hallucinations by up to 25.6% (and 14.6% on average) while preserving general accuracy. Ablations show LoRA adapters and regularization act synergistically, causal interventions prove neurodiversity as the mediating factor and correlational studies indicate scale: a 0.1% neural correlation increase is associated with a 3.8% hallucination increase. Finally, task-dependent optimality emerges: different tasks require different optimal amounts of neurodiversity. Together, our results highlight neural diversity as a third axis of scaling -- orthogonal to parameters and data -- to improve the reliability of language models at fixed budgets.

arXiv

GenGA: Editable and Data-Grounded Graphical Abstract Generation for Academic Papers

Takuro Kawada, Shunsuke Kitada, Hitoshi Iyatomi

Graphical Abstracts (GAs) visually summarize the key findings of academic papers, playing a crucial role in facilitating the understanding of research content. Recently, advancements in vision-language models and image generation models have enabled the automatic generation of scientific figures based on paper content. However, most conventional methods output the generated results as raster graphics, making post-editing (e.g., text modification and layout changes) highly difficult. This poses a significant challenge, as they are unsuitable for the iterative figure revision process inherent in paper writing and peer review. To tackle these challenges, we define the novel task of generating editable GAs from paper content and propose GenGA, a new GA generation framework that directly produces figures in vector format. By generating figures as a collection of vector elements with a hierarchical structure, GenGA produces outputs that can be seamlessly imported into existing drawing tools for intuitive, element-level editing. Furthermore, we introduce the Structural Independence Coefficient (SIC), a metric that quantifies the editing simplicity of a figure based on the degree to which local modifications propagate to other elements. Experimental results show that GenGA achieves superior editing simplicity compared to conventional methods, and even surpasses human-authored GAs in conciseness and semantic alignment. We also validate SIC as an effective metric correlated with manual editing costs. This study fundamentally redefines GA generation as an editable vector graphic generation problem grounded in the practical workflows of researchers, significantly promoting effective scientific communication.

arXiv

DBLAST: Dependent Block Drafting for Stochastic Speculative Decoding

Amirmohammad Karimi, Chao Gao, Negar Hassanpour

Speculative decoding accelerates large language models' inference by using a lightweight drafter to propose multiple future tokens and a target model to verify them. While recent block and diffusion-style drafters can predict several positions in a single pass, their training and sampling procedures are typically optimized for greedy decoding or assume that positions in the draft block are conditionally independent. This assumption becomes brittle in non-greedy speculative decoding, where the target distribution is deliberately stochastic and multiple continuations become plausible. We study this mismatch for block diffusion drafters and show that the accepted draft length degrades as the entropy of the target sampling distribution increases. We propose a dependent block drafter based on a low-rank latent mixture over token positions, complemented by an acceptance-oriented training objective that directly targets the expected verified length. Experiments with Qwen3-4B and Qwen3-8B on GSM8K, MT-Bench, HumanEval, and creative-writing benchmarks show that our approach, namely DBLast, consistently improves accepted length over independent block sampling, especially in higher-entropy decoding regimes.

arXiv

Example-Guided Prompting for Document-Level Text Simplification

Marina Litvak, Ariel Perstin, Ilan Shtilman, Michael Färber

Document-level text simplification requires large language models (LLMs) to rewrite complex documents while preserving meaning, readability, and discourse coherence. Although prompt-based LLMs have shown promising performance, they often produce inconsistent simplifications because textual instructions alone provide limited guidance for complex document-level transformations. We investigate whether retrieved document-simplification examples can improve document-level generation by augmenting prompts with examples selected from a parallel simplification corpus. This example-guided prompting approach enables LLMs to exploit relevant simplification patterns without task-specific fine-tuning. Experiments on the OneStopEnglish corpus using multiple state-of-the-art LLMs show that incorporating retrieved examples consistently improves simplification quality over prompt-only generation and achieves competitive or superior performance compared with representative supervised (T5) and planning-based (PlanSimp) document simplification systems. Furthermore, we find that the benefits of example-guided prompting vary across LLMs, suggesting that effective use of retrieved examples depends on a model's ability to integrate contextual information during generation.

arXiv

EvoHarness-RL: Learning Self-Evolving Runtime Harness for Long-Horizon LLM Agents

Xuying Ning, Dongqi Fu, Tianxin Wei, Hanqing Zeng, Yuanchen Bei, Bingxuan Li, Zihao Li, Qifan Wang, Xiang Shen, Yifan Wu, Jiayi Liu, Hong Li, Yinglong Xia, Xiangjun Fan, Hanghang Tong, Jingrui He

Long-horizon LLM agents increasingly rely on external execution support to maintain state, track progress, invoke tools, verify outcomes, and reuse experience across interactions. However, effective harness use raises two coupled challenges: state formation from noisy interaction traces and runtime control over external-state access. Existing agents usually handle both through prompts, heuristics, or domain-specific conventions, leaving the external workspace and its usage policy manually engineered. To address this, we study the problem of harness policy learning, where agents learn harness policies offline and deploy them to construct and update external harness state online during runtime task execution. We introduce EvoHarness-RL, which exposes Belief, Progress, and Experience (BPE) as policy-facing harness state. Supervised harness fine-tuning teaches the base agent the harness action space and how to construct useful external state, while cost-aware GRPO explores coordination policies to selectively read, update, and consolidate that state during long-horizon interaction. Instantiated on ALFWorld with a Qwen3-8B LLM, EvoHarness-RL reaches 96.9% success and reveals two key dynamics: harness annealing, where training internalizes recurring harness-use patterns into the model policy and shifts the agent from frequent harness calls toward selective external-state access, and harness evolution, where progress updates and experience consolidation refine the harness into a compact, task-adaptive state substrate. These results suggest that long-horizon agents benefit from trainable policies for constructing and coordinating with external harness workspaces, beyond simply adding stronger tools or larger memories.

arXiv

Mood Matters: How Syntactic Sensitivity Undermines Safety Alignment

Alina Klerings, Jannik Brinkmann, Heiner Stuckenschmidt, Simone Paolo Ponzetto

Large language models typically undergo post-training to align them with safety policies but there exist many sophisticated jailbreaks that sidestep established safeguards. For instance, prior work by Andriushchenko et al. (2025) has found that changing the grammatical tense from present to past can be enough to elicit harmful responses. In this work, we uncover a more general failure of non-imperative syntactic forms. We demonstrate that this syntactic vulnerability exists in 16 models up to 70B parameters, using behavioral evaluation. To investigate the root cause, we apply causal mediation analysis, finding that refusal is partially conditioned on upstream syntactic features. By steering these purely syntactic features we are able to trigger and suppress refusal. Finally, we trace this ill-conditioning to linguistically biased post-training data of open-source models and show that increasing syntactic diversity can mitigate the issue. Our findings suggest that current alignment approaches introduce confounders that prevent a pure semantic grounding of the refusal decision.

arXiv

The interface of intonation and lexical tone: Boundary phenomena in Mandarin varieties

Cong Zhang, Yiya Chen

This chapter explores the intricate interplay between intonation and tone in Mandarin Chinese varieties, focusing on f0, the primary acoustic cue for both intonation and tone. The main empirical base is intonation boundary phenomena, where intonation and tone intersect and influence each other in conveying a range of sentence-level linguistic functions -- such as question vs. statement -- and a rich array of speakers' attitudinal information. Theoretical models and emerging techniques are also discussed to account for the observed interactions of tonal aspects and boundary phenomena to convey multiple levels of communicative meanings.

arXiv

QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding

Ayushman Garg, Akshita Gupta, Shaswata Bhattacharya, Abhishek Gupta, Sandeep Kumar, Manoj Kumar

Autoregressive large language model inference is increasingly constrained by the memory footprint of the Key-Value (KV) cache. A dominant line of work reduces this footprint by evicting tokens that appear unimportant under attention-derived scores. However, such policies make an implicit irreversible decision: once a token is evicted, it cannot become useful again. We show that this assumption is brittle during decoding. Token and window importance drift as generated queries evolve, causing standard eviction policies to permanently discard states that later receive substantial attention under the full-cache model. To characterize this behaviour, we introduce Future Missed Mass and Global LIR, two diagnostics that measure future attention assigned to discarded states and the reactivation of historically inactive regions. We propose QEvict, a three-tier KV-cache management scheme that replaces binary retain-or-delete eviction with recoverable eviction. QEvict maintains high-confidence windows in full precision, stores intermediate windows in a quantized recoverable tier, and deletes only the lowest-confidence windows. During decoding, cumulative attention scores update window importance and when a quantized window becomes important again, it is dequantized and promoted to the full-precision. Under a fixed memory budget, this design preserves broader historical context while retaining exact full precision for the most important regions. Across long-context understanding, retrieval, and reasoning benchmarks, QEvict consistently improves over representative eviction and quantization baselines, reducing missed attention and improving information retention

arXiv

EdgeXpert: An Edge Device for Memory-Efficient LLM Inference with Mixture-of-Experts and Speculative Decoding

Sangwoo Ha, Hyunwoo Seo, Yurim Jo, Youngjin Moon, Hoi-Jun Yoo

On-device deployment of Large Language Models (LLMs) has become essential for personalized edge applications. A primary bottleneck is external memory access (EMA) in feed-forward network (FFN) layers. Speculative decoding and mixture-of-experts (MoE) are promising solutions. Speculative decoding reduces the number of decoding stages by generating multiple tokens per stage, and MoE minimizes per-stage cost through sparse expert activation. However, there is an incompatibility when combining these two techniques. We propose EdgeXpert, a software-hardware co-designed LLM accelerator that resolves this incompatibility. In the prefill stage, the prompt-wise expert reuse reformulates routing as prompt-level expert reuse rather than independent per-token expert selection. It identifies important tokens using a lightweight encoder, constructs a shared expert set from them, and routes less important tokens with a reduced expert budget to lower expert EMA. In the decode stage, depth-aware expert coalescing exploits the contextual similarity and mutual exclusivity of same-depth candidate tokens. Rather than loading the union of all required channels, EdgeXpert loads only salient channels and applies computational calibration to recover accuracy without additional memory access. Synthesized in Samsung 28nm technology at 800 MHz, EdgeXpert achieves up to 56.3% latency reduction and 44.1% energy reduction compared to prior works, while maintaining near-baseline accuracy.

arXiv

Constraint-First Reasoning: A Training-Free Protocol for Exploiting Answer-Space Constraints in Mathematical Problem Solving

Hongbo Ma, Bangji Yang, Yunqian Selina Cheng, Jiajun Fan, Hanwen Zhang, Ge Liu

Large language models can derive a plausible mathematical object yet still violate explicit requirements--for example, by omitting a modular reduction, returning a non-integer, or using the wrong encoded answer form. We introduce Constraint-First Reasoning (CFR), a training-free two-stage prompting protocol: Stage 1 extracts and summarizes constraints entailed by the problem, and Stage 2 solves while checking intermediate and final results against that summary. Routed-CFR activates the two-stage protocol only when a text-only regex router detects restrictive cues; otherwise it uses direct chain-of-thought (CoT). Across AIME, CMIMC, BRUMO, and AIMO_AMC, the method improves direct CoT on multiple backbones. We further report convention-controlled routing experiments, matched prompting baselines, problem-level paired tests, decoding robustness, constraint-quality audits, total-token accounting, and an OlympiadBench evaluation. These analyses position CFR as a targeted test-time intervention whose benefit depends on recoverable constraints and reliable Stage 1 extraction, rather than as a general-purpose replacement for mathematical reasoning.

arXiv

Analysis of Numerical Localisation in LLM Translations

Patrizia Kaye

The work of Tang et. al. (2025) on numerical translation is extended by analysing the capability of five large language models (LLMs) for the localisation of times, numbers, and dates instead of translation. Models were selected that could be loaded onto and run on commodity hardware and a baseline quality for each mode is computed, then three different strategies to improve on that accuracy were tested. In contrast to Tang et. al., it was discovered that on the tested LLMs, embedding the localisation principles into the prompt context provided a statistically significant improvement in accuracy compared to direct translation or the alternative strategies.

arXiv

TriQua: Reconciling Granularity and Context in Factuality Evaluation

Jin Liu, Steffen Thoma, Achim Rettinger

The "decompose-then-verify" paradigm for LLM factuality evaluation faces a fundamental trade-off: atomic facts, i.e., one sentence conveying one unit of information, often omit essential context, while broader statements lack the granularity needed for precise assessment. To address this, we introduce TriQua, a framework that flexibly models facts based on their complexity. Simple claims are extracted as standard triples, while complex claims are represented as hyperrelational facts by attaching auxiliary contextual qualifiers. This adaptive structure preserves the necessary context for accurate retrieval and verification without sacrificing atomicity. Furthermore, TriQua's verification process directly annotates concrete errors within specific triples and qualifiers, providing fine-grained explainability for error detection. Alongside the framework, we propose TriQuaScore to quantify the factuality of these structured fact units. Empirical evaluations show that TriQuaScore strongly aligns with human annotated factuality scores, TriQua achieves robust decomposition quality, and outperforms existing decomposition-based frameworks in evidence-based fact verification.

arXiv

AI Security Leaderboard: Methodology, Results and Minimal Standard

Jasper Timm, Lukas Struppek, Ziwei Xu, Grace Cheong, Oscar Mata, Dan Zhao, Mick Yang, Isadora De Andrade, Xiaojun Jia, Yiming Li, Samuel Bauer, Heather McIntyre, Adam Gleave, Edward Yee, Kellin Pelrine

The AI Security Leaderboard is an independent benchmark that ranks the safeguards of frontier AI models from least to most secure. It tests models against the FAR$.$AI Minimal Standard for Safeguards, which represents a minimum bar for security: meeting it does not guarantee a secure model, but failing to meet it guarantees a lack of state-of-the-art security. Version 1.0 covers severe misuse requests across chemical, biological, radiological, nuclear, and explosive (CBRNE) threats and offensive cybersecurity. In this report, we tested four leading models for universal jailbreaks in the context of this minimal standard, and found more than a hundredfold difference in security. Claude Fable 5 and GPT-5.6 Sol held against every attack we ran, with no universal jailbreak found; we estimate they would likely cost more than \$14,200 to jailbreak, if it is possible with this methodology at all. Meanwhile, we found hundreds of universal jailbreaks for Grok 4.5 and Gemini 3.1 Pro; each broke for under \$300, with universal jailbreaks in Grok's weakest domain, cybersecurity, accessible for as little as \$24. The gap is fixable: every weakness we found belongs to a known class of attack that already has a defense deployed in production models. The leaderboard will be updated on a rolling basis as new models are released, and the evaluation methodology and Minimal Standard will be periodically revised to take into account the latest capabilities and the state-of-the-art in safeguards. The leaderboard is available at leaderboard.far.ai.

arXiv

Activation-Guided Neuron Intervention to Induce Alzheimer's-Related Computational Language Phenotypes in a Large Language Model

Rui He, Ercong Nie, Hong Jiang, Iris E. Sommer, Philipp Homan, Wolfram Hinzen

Changes in spontaneous speech provide an early signal of cognitive dysfunction in Alzheimer's disease (AD) that large language models (LLMs) can detect. However, detection alone cannot establish whether the underlying model representations contribute functionally to behavior. We introduce an activation-guided intervention framework using Qwen3-8B. The framework identifies feed-forward neurons with higher activation rates for AD than control transcripts and modulates their output contributions during generation by scaling the corresponding down-projection weights. This yielded nine edited variants differing in intervention direction, magnitude, and scope. The original and edited models completed the same 12-turn neuropsychological battery, assessed through blinded human ratings and computational linguistic measures. Amplifying AD-associated neurons produced graded impairments in story recall, verbal fluency, working memory, procedural discourse, scene construction, and coreference resolution. Attenuation largely preserved performance and selectively improved several outcomes. Amplification also reduced lexical surprisal, idea density, syntactic complexity, and discourse quantity, broadly paralleling changes reported in human AD speech. These findings show that neurons identified solely from clinical language differences can influence behavior across multiple cognitive domains, providing proof of concept for an AD-related computational phenotype and a controlled framework for experimentally examining links between language and broader cognitive dysfunction.

arXiv

Right Knowledge, Wrong Answer: Characterizing Parametric Temporal Conflict in Open-Weight Language Models

Elias Hossain, Sourav Saha, Tasfia Nuzhat Ornee, Sanjeda Sara Jennifer, Umesh Chandra Biswas, Shubhashis Roy Dipta, Rajib Rana, Niloofar Yousefi

Language models may encode both outdated facts and their newer replacements. We introduce Parametric Temporal Conflict (PTC), where the newer fact is present and recoverable, but the default forward pass prefers the outdated one. We release a deterministically verified benchmark of 8,746 Wikidata position-holder transitions and evaluate four open-weight language models across three families. A date-prefix prompt recovers the newer fact in 61-81% of PTC cases. Activation patching flips predictions in 72-85% of cases and localizes the preference to model-specific upper-layer regions. Residual-stream steering outperforms norm-matched random directions, indicating direction-specific representations. These results show that PTC reflects a localized representational preference rather than missing knowledge. Recovery is measured on oracle-identified conflicts because automatic detection remains unreliable. We release the benchmark, code, and statistics.

Keep browsing

More papers from top AI conferences

30+ top conferences, by conference, year, and topic.