The paper extends mechanistic interpretability of large language models by modeling concepts as low‑dimensional non‑linear manifolds rather than linear subspaces. It introduces a concept‑based alignment (CBA) score to compare these manifolds across layers and models, revealing block structures in intermediate layers, a shift from syntax‑dominated to mixed syntactic‑semantic concepts, and training‑dependent multilingual sharing. The study also shows that alignment patterns differ across model families and training stages, with adjacent stages aligning more closely than distant ones.
arXiv:2510.01030v2 Announce Type: replace Abstract: The human ability to translate diverse perceptual and linguistic inputs into structured behavior has been thought to rest on learning robust repres...
arXiv:2606. 09653v1 Announce Type: new Abstract: Learned representations across models and modalities often exhibit striking structural similarities, suggesting shared underlying concept decompositions.
arXiv:2512. 07355v2 Announce Type: replace Abstract: Two traditions of interpretability have evolved side by side but seldom spoken to each other: Concept Bottleneck Models (CBMs), which prescribe what a concept should be, and Sparse Autoencoders (SAEs), which discover what concepts emerge.
The study investigates whether multilingual language models solve mathematical problems using shared or language‑specific internal representations. By replaying reasoning traces across six languages and applying Centered Kernel Alignment, the authors identify layers with cross‑language alignment and train two sparse autoencoders— a baseline and a new Geometry‑Invariant SAE (GI‑SAE) that adds an InfoNCE loss to encourage language‑agnostic activations. Functional interchangeability tests reveal that higher geometric similarity does not always translate to interchangeable features, and cross‑language sharing varies by model and depth, with GI‑SAE amplifying existing structure in some models but not others.
arXiv:2603.18908v5 Announce Type: replace Abstract: Independently trained language models often learn compatible late-stage representations, despite differences in training objectives, architectures,...
arXiv:2606. 01060v1 Announce Type: cross Abstract: Preference alignment has substantially improved the observable behavior of large language models, yet it remains unclear what alignment changes internally.
The study investigates whether monolingual language models, trained without joint multilingual objectives, develop cross-lingual alignment. By evaluating models such as Goldfish and independently built monolingual systems, the authors find that alignable representational geometry emerges across layers, strengthening with larger data, larger models, or closer linguistic proximity. A single Procrustes rotation on parallel sentences can map hidden states between models, and applying this rotation to a German model’s residuals swaps factual predictions to those of the donor English model, demonstrating functional transfer.
arXiv:2609.23065v1 Announce Type: cross Abstract: Chain-of-thought (CoT) can sound plausible yet be unfaithful to the model's underlying reasoning. Most prior work probes CoT faithfulness through inp...
arXiv:2606. 09850v1 Announce Type: new Abstract: Post-training alignment algorithms are predominantly evaluated as black boxes, obscuring how they reshape language models' internal computations.
arXiv:2602. 07026v3 Announce Type: replace-cross Abstract: Despite the success of multimodal contrastive learning in aligning visual and linguistic representations, a persistent geometric anomaly, the Modality Gap, remains: embeddings of distinct modalities expressing identical semantics occupy systematically offset regions.
The study compares an English-only and a bilingual decoder-only model, each 310 M parameters, trained on eight diverse languages while controlling for English exposure, compute, and document overlap. After aligning on shared English vocabulary, the authors find that token embeddings appear similar, but the deeper hidden states used for prediction diverge across models. This hidden‑state mismatch grows through middle transformer layers and persists despite controls, indicating that contextual processing differs between the models. "whyItMatters":"The findings show that embedding alignment can conceal significant internal representation differences, which is crucial for any downstream work that assumes aligned multilingual models are interchangeable."
arXiv:2608. 06300v1 Announce Type: new Abstract: Automatic speaking assessment systems are increasingly deployed in high-stakes settings to mark second language (L2) learners' speaking tests, making it critical to show that their scores depend on speaking proficiency rather than irrelevant speaker attributes such as first language (L1) or age.