artifact_id stringlengths 7 41 | category stringclasses 4
values | description stringlengths 24 68 | format stringclasses 1
value | path stringlengths 20 52 | sanitized bool 2
classes | sha256 stringlengths 64 64 | size_bytes int64 1.01k 384k |
|---|---|---|---|---|---|---|---|
activation_analysis | summary | Activation saturation and bounding analysis summaries. | json | results/activation_analysis.json | false | d9b7e962cf7f075119968651e0a856ca73b7228a1a283f8b68d7d4fda3cdd232 | 1,618 |
alpha_sweep | summary | Alpha-sweep summary for DyT bounding strength controls. | json | results/alpha_sweep.json | false | 038699ec0852c48221aa62796c998c38928bddba245fe12386799634425bbf0e | 1,014 |
convergence_10k | summary | Short-horizon convergence summary used for consistency checks. | json | results/convergence_10k.json | false | 9d721c4ff939d5955e5b886235c4fd1841b3c3e6bf61b40440a5a6d7474f6942 | 1,740 |
full__activation_rank_v1__aggregate | aggregate | Aggregate metrics for activation rank v1. | json | results/full/activation_rank_v1/aggregate.json | false | 9d92395f143c52de2e2fa40e593f0972fd75e044148ea822e062902f1d538531 | 69,999 |
full__all_results | full_result | Sanitized aggregate table across public result cells. | json | results/full/all_results.json | true | 7797a498acfb8f2803bdc2d062e9550f4347d33b298864c59a6f08176b31a88e | 116,748 |
full__crossover_predictor | full_result | Predictor-validation summary for regime sign checks. | json | results/full/crossover_predictor.json | false | 4ac44ce544860ab9eb771c2defc97917064311985ed71e204751a62a285550a4 | 6,370 |
full__downstream_v5__aggregate | aggregate | Aggregate metrics for downstream v5. | json | results/full/downstream_v5/aggregate.json | false | 974f3620ec33957cdc6afc8ebd0146b62e8874dc41136bf5b9b6c3bdd7d30ff6 | 4,897 |
full__fast_extract | full_result | Fast extraction summary for public artifact consistency checks. | json | results/full/fast_extract.json | true | 5881efea1ede8f29296fb7a2108f64e5a8de0f548c5c784056bff1fb46a65e93 | 15,815 |
full__hessian_pyhessian_v4__aggregate | aggregate | Aggregate metrics for hessian pyhessian v4. | json | results/full/hessian_pyhessian_v4/aggregate.json | false | 3b887dccad0b445a0bf74ff84ddc42dcac1417a19f6a117994449fd0de48f60e | 22,913 |
full__lambada_eval__aggregate | aggregate | Aggregate metrics for lambada eval. | json | results/full/lambada_eval/aggregate.json | true | ce37d10cf3456d3fda311ddc97ff32f377b5272f81d9ec80e5bd6dfed37d8b61 | 27,980 |
full__manifests__alpha_sweep_118m_summary | manifest | Manifest for the 118M alpha-sweep control. | json | results/full/manifests/alpha_sweep_118m_summary.json | true | ecd533b8152d2b6e348b4f94ff487fad561cb2a552d25a5ac557223832a52689 | 2,281 |
full__manifests__predictor_validation | manifest | Manifest for predictor-validation outputs. | json | results/full/manifests/predictor_validation.json | false | 14302e82266cfe5b39cfd6c0ea755a57da256a0c41f51f027f6a6183e753ab51 | 16,400 |
full__manifests__sig_tests | manifest | Statistical-test manifest for reported comparisons. | json | results/full/manifests/sig_tests.json | false | 9d10175c33f9c316f97f4039cb4c9bb524459e922c017ba28c1e23ffce8990f5 | 11,895 |
full__noise_stability_s3__aggregate | aggregate | Aggregate metrics for noise stability s3. | json | results/full/noise_stability_s3/aggregate.json | false | f5d601c707aa63e4c234ba59abc14e46baa4d4998475eb245ecfbd12e495aba7 | 3,127 |
full__noise_stability_uniform__aggregate | aggregate | Aggregate metrics for noise stability uniform. | json | results/full/noise_stability_uniform/aggregate.json | false | 3ac8a1f87929bfea98e19cbf7ba8be4c4fca769321ae4f7024faa7a68be6cf7f | 15,271 |
full__saturation_results | full_result | Saturation diagnostic outputs across calibration cells. | json | results/full/saturation_results.json | false | bb871cb3e3c7cb95e8b334d042fb9cbfb5d3b0ed4bf45c6e3501cb3ef255c55a | 384,041 |
full__scale5_metadata_check | full_result | Scale-5 metadata consistency summary. | json | results/full/scale5_metadata_check.json | false | 0127598467a2bb63261d2587144c2101bd58ac12330c2cff0fa3416f379d6c53 | 13,138 |
phase_diagram | summary | Regime-level phase-diagram summary across data and scale. | json | results/phase_diagram.json | false | ac6ee3bde5369f02c23961aa2b27e3572700526c3a5e06ba96ce55b779dfba68 | 2,130 |
rmsnorm_results | summary | RMSNorm comparison summary. | json | results/rmsnorm_results.json | false | fec2c4f67bef0c77485cea5c42258952c056ba252b89b9957c49867126e48aee | 1,242 |
scaling | summary | Scaling summary across reported model/data regimes. | json | results/scaling.json | false | f900f06124ba36f95b3c801db285c6e7e4c8331322687067ace79b50d5437c05 | 3,247 |
train_val_gap | summary | Train/validation gap summary used for regularization interpretation. | json | results/train_val_gap.json | false | c46d32f450711887a3dfca4f13ab468c4e4a3d0f88b4198c608365d09cf10e4d | 2,026 |
vit_results | summary | ViT cross-check summary. | json | results/vit_results.json | false | af2028744c89e4af0c40b4df861cc6ebe923f2193693c15de1181fdd5f14acc4 | 1,975 |
DyT Composition Study Artifacts
This dataset contains sanitized result manifests and analysis outputs for When Does Removing LayerNorm Help? Activation Bounding as a Regime-Dependent Implicit Regularizer.
DOI: https://doi.org/10.48550/arXiv.2604.23434
Contents
The artifacts include aggregate training metrics, saturation measurements, statistical-test summaries, predictor-validation outputs, table-source manifests, and selected aggregate analysis files used by the public code repository.
The Dataset Viewer table is an index of the artifact files. The machine-readable result artifacts are stored under results/.
This is not a natural-language training dataset. It does not redistribute WikiText, OpenWebText, LAMBADA, BLIMP, model checkpoints, or raw training logs.
Intended Use
Use this artifact bundle to:
- inspect the machine-readable results behind the paper;
- reproduce paper tables and consistency checks;
- compare DyT, LayerNorm, RMSNorm, HardTanh, DiffAttn, and related controls at the reported scales;
- audit provenance for reported quantitative claims.
Limitations
- The experiments are compute-limited and below Chinchilla-optimal training.
- The included files are result artifacts, not full raw training traces or checkpoints.
- The saturation diagnostic should be treated as a per-deployment calibration cue, not a universal rule.
- Raw public datasets retain their original licenses and are not mirrored here.
Licensing
The result artifacts in this dataset are released under CC BY 4.0.
The associated GitHub code is released under the MIT License.
Citation
@misc{verma2026dytcomposition,
title = {When Does Removing LayerNorm Help? Activation Bounding as a Regime-Dependent Implicit Regularizer},
author = {Verma, Lucky},
year = {2026},
publisher = {arXiv},
doi = {10.48550/arXiv.2604.23434},
url = {https://arxiv.org/abs/2604.23434},
eprint = {2604.23434},
archivePrefix = {arXiv},
primaryClass = {cs.LG}
}
- Downloads last month
- 35