arXiv Machine Learning
Aug 5
Wiring Beats Blending: What Transfers Between Transformer Sizes -- and What Doesn't
arXiv:2608. 02829v1 Announce Type: new Abstract: Model families train every size from scratch.
Wiring Beats Blending: Structure-Aware Compensation for Transformer Downscaling
The paper investigates converting a large pretrained transformer (1.4 B parameters) into a smaller sibling (410 M) by studying representation alignment and parameter projection. It finds that dense weight projection destroys structure, and that a low‑budget, structure‑aware compensation—separating least‑squares function alignment from variance‑preserving rescaling—yields significant gains on token‑efficient training, outperforming subcloning and standard distillation pipelines at matched budgets.
Related stories
arXiv Machine Learning
Sep 3
Train What You Deploy: Closing the MLP Reachability Gap in Low-Rank Clone Distillation
The paper introduces a new approach to low‑rank clone distillation that ensures the student model’s training targets exactly the weights it will use at inference. By redefining the training objective to cover the full deployed matrix—without changing the model’s shape, parameter count, or FLOPs—the authors recover previously unreachable linear degrees of freedom. This results in significant performance gains across multiple teacher models, achieving comparable or superior accuracy with fewer tokens and parameters.
arXiv AI
Sep 11
Scaling Post-Training Ternarisation to Qwen3-8B Capability Retention, Reproduction, Lossless Packing, and Packed Execution
The paper reports a large‑scale post‑training ternarisation of the Qwen3 language model, extending a conversion pipeline from the 4B to the 8B variant. Using KOTMS rotation, E2M‑ATQ adaptive ternarisation, and GPTQ‑style error compensation, the authors achieve a 1.361× perplexity ratio across three corpora and retain 78.5% of the FP16 accuracy on zero‑shot tasks, with the 8B model outperforming the 4B by 8.9 percentage points. The study also demonstrates lossless lattice‑aware packing, producing an 8.24 GiB checkpoint that preserves perplexity, and shows that direct packed execution can reach 15.52 tokens/s in 7.35 GiB, though packed GEMV remains slower than FP16 cuBLAS.
arXiv AI
Sep 10
Everything in Moderation: Per-Domain Coverage Optima and Alignment-Resistant Domain Gaps in Multi-Domain Mid-Training
The study investigates how the composition of data during the mid‑training phase of language models affects performance across multiple domains. Experiments with Qwen3‑8B‑Base on five distinct KOR‑Bench domains show that moderate coverage (10%‑40%) yields the best per‑domain results, and that alignment passes cannot fully close the performance gaps created by mid‑training data choices. Additionally, zero coverage in mid‑training severely degrades accuracy, while a carefully tuned allocation can provide the largest overall pipeline improvement.
arXiv AI
Jul 20
Kolmogorov--Arnold Networks for Small Language Models
arXiv:2607. 15525v1 Announce Type: cross Abstract: Kolmogorov--Arnold Networks (KANs) replace fixed node activations with learned one-dimensional edge functions, offering an explicit interface for interpretation and a possible alternative to transformer feed-forward networks.
arXiv AI
Sep 15
Depth and Scale in the Sub-150M Regime: JugnuLM-53M vs JugnuLM-110M
arXiv:2609.14715v1 Announce Type: new Abstract: We scale our conventional sub-150M pretraining recipe from 53.5M to 109.7M parameters, holding the method fixed (Qwen3-style decoder with grouped-query...
arXiv Machine Learning
Sep 17
Beyond Embedding Transfer: Component Roles in Grokking Transfer and Stability
The study investigates which components of a neural network contribute to rapid generalization (grokking) and how stable that improvement remains during further training. By transferring internal attention and MLP weights along with token embeddings and readout, the authors achieve a 5.46‑percentage‑point boost in early accuracy and a 558‑step reduction in confirmation latency, while also demonstrating that freezing transferred representations largely prevents post‑grokking relapse. The work delineates clear component‑level differences between acceleration and stability, and identifies architectural limits where omitting donor embeddings leads to significant performance loss.
arXiv AI
Jun 2
Lying Is Just a Phase: The Hidden Alignment Transition in Language Model Scaling
arXiv:2605. 18838v3 Announce Type: replace-cross Abstract: Scaling laws predict loss from compute but not how capabilities interact.
arXiv Machine Learning
Jun 3
Pruning Deep Neural Networks via the Marchenko--Pastur Distribution
arXiv:2606. 02608v1 Announce Type: new Abstract: We study a Marchenko--Pastur (MP) random-matrix approach to pruning deep neural networks with very small post-pruning fine-tuning budgets.
arXiv Machine Learning
Jul 17
Adaptive Runge-Kutta Step Control Buys Training Loss, Not Generalization: An Honest Compute-Matched Study of RK-Adam Optimizers
arXiv:2607. 14516v1 Announce Type: new Abstract: Interpreting optimizers as gradient-flow discretizations has motivated applying higher-order Runge-Kutta (RK) integrators to neural networks.
arXiv Machine Learning
Jul 16
Weight Feedback Computes the Jacobian Transpose Locally in Modern Deep Networks
arXiv:2607. 13380v1 Announce Type: new Abstract: Predictive Coding (PC) offers a biologically motivated alternative to backpropagation via local weight updates, yet routing error between layers still relies on an autograd Jacobian-transpose ($J^\top$) product - the last non-local operation in PC.
Hugging Face Trending Papers
Sep 8
Everything in Moderation: Per-Domain Coverage Optima and Alignment-Resistant Domain Gaps in Multi-Domain Mid-Training
The paper investigates how per-domain data composition during the mid‑training phase (between pre‑training and alignment) affects model performance. Experiments with Qwen3‑8B‑Base across five KOR‑Bench domains show that a moderate coverage band (10%‑40%) yields the best performance for each domain, and that alignment passes cannot fully close the gaps created by suboptimal mid‑training allocations. Additionally, zero coverage during mid‑training severely degrades accuracy, while a carefully tuned allocation can provide the largest overall pipeline improvement.