6 papers
When Dynamic Scaling Fails: A Scale-State Feedback Governs Recurrent-State Quantization in Linear-Attention and SSM LLMs
Yuhang Jiang
Linear-attention and state-space LLMs carry long context in a fixed-size recurrent state, and quantizing that state is the natural way to cut its cost during long decode. Unlike a…
Chaos Is a LADDER: Domain Generalization Beyond Invariance via Reweighting
Yuhang Jiang, Fengchuan Zhang, Sanguo Zhang +1
Domain generalization (DG) aims to learn from multiple source domains and generalize to unseen target domains. Most DG methods pursue invariance: they seek a causal representation…
A Circuit, Not The Circuit: Non-Unique Causal Localisation of the Mamba-2 State Sink
Yuhang Jiang, Bowen Zhang
Mechanistic interpretability routinely reads a probe and labels its top-activating units as the circuit executing the computation. We test the move in Mamba, on the state sink: the…
Refit the Probe: Single-Direction Ablation Is Not a Necessity Test
Yuhang Jiang
Probes are routinely paired with an intervention: ablate the direction the probe found, run the model, and read the change in task accuracy, taking a large drop as evidence that th…
PInVerify: An Offline Embodied Benchmark for Active Instance Verification
Yuhang Jiang
Embodied agents have made strong progress in navigating to target objects, but reaching the goal vicinity does not guarantee that the agent has found the correct instance: subtle a…
Can ChatGPT Overcome Behavioral Biases in the Financial Sector? Classify-and-Rethink: Multi-Step Zero-Shot Reasoning in the Gold Investment
Shuoling Liu, Gaoguo Jia, Yuhang Jiang +2
Large Language Models (LLMs) have achieved remarkable success recently, displaying exceptional capabilities in creating understandable and organized text. These LLMs have been util…