arXiv:2604. 13082v2 Announce Type: replace-cross Abstract: Grokking in transformers trained on algorithmic tasks is characterized by a long delay between training-set fit and abrupt generalization, but the source of that delay remains poorly understood.
arXiv:2607. 20652v1 Announce Type: cross Abstract: Language models are thought to exhibit the phenomenon of superposition, representing many more features than dimensions in their residual streams.
arXiv:2602. 22600v2 Announce Type: replace-cross Abstract: Training selects for behavior, not circuitry: many weight configurations can implement the same function.
Scaling Large Language Models (LLMs) has been driven mainly by enlarging the Transformer backbone, but for an already-strong model this requires another round of costly pretraining. We study whether an existing backbone can keep improving by allocating more computation to each token while leaving the Transformer backbone fixed.
arXiv:2607. 04333v1 Announce Type: new Abstract: Grokking -- generalization arriving long after training-set interpolation -- can be accelerated by structure-agnostic interventions: gradient filtering, weight-norm clamping, geometric penalties on hidden representations.
arXiv:2606. 26488v1 Announce Type: new Abstract: Recursive reasoning models can solve complex structured tasks with only a few million parameters by repeatedly updating a latent state.
arXiv:2601. 22002v5 Announce Type: replace Abstract: Transformers achieve superior performance on many tasks, but impose heavy compute and memory requirements during inference.
arXiv:2510. 04212v4 Announce Type: replace-cross Abstract: The pursuit of computational efficiency has driven the adoption of low-precision formats for training transformer models.
arXiv:2607. 22361v1 Announce Type: new Abstract: We study information bottlenecks in modern deep-learning architectures -- RNNs, softmax transformers, linear-attention transformers and state-space models -- through the lens of the indexing primitive.
arXiv:2512. 18934v2 Announce Type: replace-cross Abstract: Catastrophic forgetting poses a fundamental challenge in continual learning, particularly when models are quantized for deployment efficiency.
arXiv:2605. 18324v2 Announce Type: replace-cross Abstract: Representation Autoencoders (RAE) replace traditional VAE with pretrained vision encoders.
arXiv:2608. 03461v1 Announce Type: new Abstract: Decomposition-based Programming-by-example (PBE) scales performance by splitting tasks into subtasks that a learned synthesizer solves: a decomposer predicts intermediate subgoals, and a synthesizer generates programs conditioned on them.
arXiv:2608. 12385v2 Announce Type: replace Abstract: As large language models serve ever more requests, cumulative inference cost is growing relative to the one-time cost of training.