This work proposes extracting mechanism mounts directly from linear sites by column-tiled SVD: each mount is a triple read as trigger, write, and strength, and evaluates mounts with a pre-registered suite judged on full-write energy lift rather than tile-local lift.
Abstract
The dominant approach to mechanistic interpretability trains proxy dictionaries such as sparse autoencoders and labels features from max-activating text. The best such atlases identify con- cepts, but that identity lives in the learned dictionary rather than in the network weights them- selves. We propose extracting mechanism mounts directly from linear sites by column-tiled SVD: each mount is a triple (v,u,{\sigma}) read as trigger, write, and strength. Identity is the weight rule. We evaluate mounts with a pre-registered suite judged on full-write energy lift rather than tile-local lift. On Gemma-2-2B with WikiText-2 (16,384-token subsample), all seven linear maps are scored: residual writes (mlp.down, attn.o) receive full A/B/C with steer after post-sublayer RMSNorm and pass 52/52 site-layers; other maps receive A/B only (mlp.gate/attn.q/attn.k/effective mlp.up/attn.v 26/26 each). Aggregate: 182/182 GO. We release library code, the corpus builder, the experiment entrypoint, and unit tests.
A family of difference-informed pruning methods built upon this principle are introduced, suggesting that preserving output differences is a broadly useful and composable signal for post-training LLM sparsification.
Linghao Kong, Inimai Subramanian, Micah Adler et al.· 0 citations
PoLoRA is introduced, a Preconditioned Orthogonalized LoRA optimizer built from three ingredients: a product-aware spectral update direction, curvature preconditioning derived from controlling the per-sample loss change, and a magnitude rule that controls the sizes of both the factor and merged updates.
Nikhil Ghosh, Tetiana Parshakova, Robert M. Gower· arXiv.org· 1 citation
Language models are thought to exhibit the phenomenon of superposition, representing many more features than dimensions in their residual streams. Sparse autoencoders (SAEs) are designed to recover such features post-hoc, but training models that are interpretable by construction has remained impractical, as a per-layer over-complete bottleneck is prohibitively expensive in both memory and compute. To overcome this issue, we introduce the ParityTransformer, a GPT-2-scale architecture whose intermediate representations are efficient and wide / sparse by design. At each layer, a Deep Parity Bottleneck (DPB) replaces a learned over-complete basis with a parameter-free algebraic dictionary, providing a deterministic incoherence guarantee and eliminating the memory requirements that have prevented per-layer interpretable bottlenecks at scale. A DPB is a hierarchically structured sparse bottleneck which efficiently enforces sparsity using a multi-level mixture-of-experts approach: a hardware-aware implementation that closes the cost gap between activation sparse and dense training to a manageable interpretability tax. Empirically, ParityTransformers perform at least as well as post-hoc SAEs on sparse probing tasks, while out-performing on measures of feature absorption, steering effectiveness, and fine-grained causal interventions. Because subsequent computation acts only on features that survive the sparse bottleneck, the ParityTransformer's features are native to the model's forwards pass by construction, addressing the question of whether SAEs probe features the model actually uses during computation. We see this as a step toward training models whose internal representations are interpretable by design rather than recovered post hoc.
Andrew Mack, K. Tou, Mark C. Henry et al.· arXiv.org· 0 citations
F-WANDA is introduced, a drop-in modification of WANDA that reallocates the per-row keep budget across output neurons in proportion to the empirical Fisher information of the pre-activation, placing F-WANDA on the Pareto frontier of quality versus pruning cost for sustainable LLM compression.
This work introduces Giga-Embeddings, a family of text embedding models designed to combine strong retrieval quality with efficient serving, and trains the compact model using a dimension-agnostic objective that aligns teacher and student similarity distributions.
Egor Kolodin, Egor Krasnoperov, Evgeniy Kosarev et al.· 0 citations
Equivariant networks are commonly trained with Adam, yet recent work reports that matrix-structured optimizers such as Muon can perform better on these architectures without explaining why. We identify one source of this difference inside equivariant linear layers. Each irrep block learns a channel-mixing matrix $W_l$ shared across its $2l+1$ components, giving the expanded map $W_l \otimes I_{2l+1}$. For a single application of the layer, the gradient of $W_l$ sums $2l+1$ outer product contributions and has rank at most $2l+1$. Adam rescales stored weights individually without using the irrep boundaries, so one learning rate can produce different spectral step sizes across blocks within a layer. We address this mismatch by normalizing each block update separately, without introducing a new hyperparameter. This changes only the scale of the update, leaving Adam's moment estimates and its direction within each block unchanged. We evaluate the mechanism in a controlled $\mathrm{SO}(3)$-equivariant model with a matched dense control and in an e3nn interatomic potential model trained on rMD17 and MD22. The toy setup isolates a mismatch that grows with width while the dense control shows no corresponding growth. In the interatomic potential model, block normalization and tuning Adam's momentum coefficients independently improve performance, but neither alone matches Muon. Combined, they make Adam competitive with Muon on all datasets, indicating that blockwise step control and momentum accumulation account for much of Muon's advantage.
Andrei Manolache, Mathias Niepert· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
