arXiv:2608. 03842v1 Announce Type: cross Abstract: When a language model fails on surface-perturbed input (typos, OCR noise, homophones), "which layer is responsible" has three natural operationalizations: where representations diverge most (sensitivity), where restoring clean activations recovers the prediction (causality), and where a small adapter can repair the damage (compensatory capacity) - and we show these three layer maps dissociate.
By Nathan Labiosa, David Buff, Ena Nayak, Erica Donno
arXiv:2607. 19618v1 Announce Type: cross Abstract: Genomic language models achieve strong performance across regulatory-genomics tasks, yet what these models internally represent remains opaque, and the field lacks a principled procedure for verifying that an apparent ``concept'' inside a model is real rather than an artifact of sequence composition.
By Sarwan Ali
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.
By Fernando Cardenas Piepereit
arXiv:2601. 18699v2 Announce Type: replace Abstract: Sequential fine-tuning of Large Language Models (LLMs) adaptation to target tasks often triggers catastrophic forgetting, where the acquisition of novel target skills degrades ancestral capabilities.
By Gustav Olaf Yunus Laitinen-Fredriksson Lundstrom-Imanov
arXiv:2608.12149v2 Announce Type: replace
Abstract: We present the first systematic study of Massive activations (MAs) in layer-interleaved HLA LLMs and uncover two architecture-aligned morphologies:...
By Zunhai Su, Bohan Sun, Xialie Zhuang, Shuibai Zhang, He Xiao, Jing Xiong, Hengyuan Zhang, Zhongzhu Zhou, Tiantian Zhang, Ngai Wong, Chuan-Wei Kuo
We present the first systematic study of Massive activations (MAs) in layer-interleaved HLA LLMs and uncover two architecture-aligned morphologies: MAs consistently spike immediately before full attention layers, forming pre-attention spikes (PAS), and can persist through intervening linear attention layers, giving rise to inter-spike plateaus (ISP). As full attention becomes denser, successive PAS become increasingly connected through ISP, ultimately recovering the stable MA morphology of full attention LLMs.
The paper investigates how six naturalistic and synthetic input perturbations affect decoder‑only language models at three levels: output behavior, hidden‑state geometry, and attention‑head function. Using GPT‑2 and Qwen2.5 checkpoints, the authors analyze layerwise geometry with centered kernel alignment and intrinsic dimension, and examine attention‑head responses in GPT‑2. They find that perturbation types produce distinct metric profiles that are not fully captured by output measures and vary across checkpoints, highlighting the need for multi‑level evaluation of robustness.
By Dun Li Chan, Emily Liu, Niyathi Allu, Christian Hoang
arXiv:2608. 12334v1 Announce Type: cross Abstract: Despite the impressive multilingual capabilities of Large Language Models, the latent dynamics dictating language selection remain poorly understood.
By Arnav Srivastav
arXiv:2607. 20524v1 Announce Type: new Abstract: Mean cross-positional attention degradation is widely reported in transformer interpretability, yet whether it causally limits contextual retrieval remains untested.
By Sagar Dangal, Manoj Shakya
The study investigates how post‑training of large autoregressive language models (ARMs) into masked diffusion models (MDMs) affects their internal computation. Across two 7B ARM‑MDM families and four diagnostic tasks, the authors find that MDMs retain much of the ARM’s high‑attribution pathways on prefix‑dominant tasks, but reorganize computation toward earlier layers on globally constrained tasks. Component‑level probes reveal that ARMs depend on sharply specialized components, whereas MDMs show weaker specialization and more diffuse output‑space alignment.
By Injin Kong, Hyoungjoon Lee, Yohan Jo
arXiv:2608. 05164v1 Announce Type: cross Abstract: Independently trained large language models may develop shared internal representations of semantic concepts despite architectural differences -- but whether this geometric similarity has functional consequences for cross-model behavioural control remains untested.
By Ayushi Agarwal
arXiv:2607. 20596v1 Announce Type: new Abstract: Sparse autoencoder (SAE) features are used to interpret and steer large language models, yet whether a feature's causal role is stable across SAE families remains untested.
By Seonglae Cho, Zekun Wu, Kleyton Da Costa, Rishi Kalra, Ilham Wicaksono, Adriano Koshiyama