IHEval: Evaluating Language Models on Following the Instruction Hierarchy

IHEval evaluates language models on instruction hierarchy compliance, finding they frequently disregard system-level priority instructions.

Zhihan Zhang, Shiyang Li, Zixuan Zhang, Xin Liu and 10 more

Published 2025 · 4 citations

Poison-then-Hide: Finetuning-Activated Backdoor Attack on Pretrained Vision Encoders

Qixuan Jin, Abinitha Gourabathina, Vinith Suriyakumar, Walter Gerych and 1 more

Paris Poster Session 2, Wed, Dec 9, 5:00 PM–7:00 PM, Paris Poster Hall · Published 2026

Evaluating and Understanding Scheming Propensity in LLM Agents

Realistic agent settings show minimal scheming despite high incentives, with model-organism scheming brittle to tool removal and oversight.

Mia Hopman, Jannes Elstner, Maria Avramidou, Amritanshu Prasad and 1 more

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

On the Depth of Monotone ReLU Neural Networks and ICNNs

Monotone ReLU networks cannot compute or approximate maximum, ICNNs need depth n for it, and depth-k ICNNs cannot simulate some depth-2 ReLU networks.

Egor Bakaev, Florestan Brunck, Christoph Hertrich, Daniel Reichman and 1 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Hybrid-LoRA: Bridging Full Fine-Tuning and Low-Rank Adaptation for Post-Training

Hybrid-LoRA selectively applies full fine-tuning to modules poorly suited to low-rank adaptation and uses LoRA elsewhere, matching full fine-tuning performance with 10% of modules fully tuned and outperforming PEFT baselines by 4.36%.

Chengqian Zhang, Wei Zhu, Kyumin Lee

Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026