arXiv:2606. 19946v1 Announce Type: cross Abstract: Activation steering controls model behavior by modifying intermediate hidden states at inference time without retraining.
arXiv:2606. 01923v1 Announce Type: cross Abstract: Large Language Models (LLMs) frequently exhibit "contextual disregard" when faced with input evidence that conflicts with their internal parametric memory, leading to persistent factual hallucinations.
arXiv:2605. 28854v2 Announce Type: replace-cross Abstract: Large language models (LLMs) exhibit remarkable flexibility in adapting to novel tasks from in-context examples without parameter updates, a capability known as in-context learning (ICL).
arXiv:2606. 27786v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) enhances LLMs by incorporating external knowledge to support response generation.
arXiv:2507. 18043v2 Announce Type: replace-cross Abstract: Inference-time steering methods offer a lightweight alternative to fine-tuning large language models (LLMs) and vision-language models (VLMs) by modifying internal activations at test time without updating model weights.
arXiv:2607. 08839v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) are typically designed under the assumption that all modalities available during training will also be accessible at inference.
MetaSteer is a new method for steering large language models that learns nonlinear, context-dependent interventions applied to attention projection matrices. Unlike traditional linear, context-independent techniques, MetaSteer adapts its effects based on the input, requiring no linear concept-geometry assumption. Trained once on a pooled preference corpus, it transfers zero‑shot to unseen concepts and out‑of‑distribution contexts, matching or surpassing strong task‑specific baselines on multiple benchmarks and model families.
arXiv:2608. 10537v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) have helped uncover mechanistic explanations for LLM behaviours such as reasoning, jailbreaking etc.
The paper audits frozen decoder‑only large language models (LLMs) on geometric reasoning tasks using parametric CAD constraints. It probes hidden states for linear decodability, forced‑choice generation, activation‑level influence, and behavioral steerability, finding that pretraining improves decoding of local geometric relations but not sketch‑level DOF status. The study shows that decodable information is not always actionable: generation often fails to express it, and steering interventions do not reliably control outputs, revealing divergences among decodability, generation, activation influence, and steerability.
arXiv:2609.36337v1 Announce Type: new Abstract: Tabular foundation models achieve strong performance by conditioning on labelled examples in context, but softmax attention limits their use on large d...
Large language models increasingly operate as tool-using agents, where small format, argument, or function-call errors can invalidate otherwise plausible responses. We study inference-time feed-forward network (FFN) intervention for improving structured outputs without retraining model weights.
arXiv:2608. 16347v1 Announce Type: cross Abstract: Direct communication between AI systems relies on natural language as an intermediate layer, incurring encoding/decoding overhead, token cost, and latency.
The paper introduces Where to Steer (W2S), a framework that adaptively selects which layer of a large language model to intervene on based on the input. It argues that optimal steering layers vary across inputs and demonstrates that W2S outperforms fixed-layer steering methods on several LLMs and alignment tasks, both in‑distribution and out‑of‑distribution. The study highlights input‑dependent layer selection as a complementary design dimension to steering vector extraction.