{"paper":{"title":"HyperNetworks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A hypernetwork generates the weights for another network to enable non-shared weights in LSTMs.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrew Dai, David Ha, Quoc V. Le","submitted_at":"2016-09-27T05:57:00Z","abstract_excerpt":"This work explores hypernetworks: an approach of using a one network, also known as a hypernetwork, to generate the weights for another network. Hypernetworks provide an abstraction that is similar to what is found in nature: the relationship between a genotype - the hypernetwork - and a phenotype - the main network. Though they are also reminiscent of HyperNEAT in evolution, our hypernetworks are trained end-to-end with backpropagation and thus are usually faster. The focus of this work is to make hypernetworks useful for deep convolutional networks and long recurrent networks, where hypernet"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our main result is that hypernetworks can generate non-shared weights for LSTM and achieve near state-of-the-art results on a variety of sequence modelling tasks including character-level language modelling, handwriting generation and neural machine translation, challenging the weight-sharing paradigm for recurrent networks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the hypernetwork can be effectively trained end-to-end with backpropagation to produce high-quality weights for the main network without introducing instability, overfitting, or requiring excessive additional computation.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Hypernetworks generate weights for a main network, allowing LSTMs to use non-shared weights and achieve near state-of-the-art results on sequence modeling tasks while using fewer parameters overall.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A hypernetwork generates the weights for another network to enable non-shared weights in LSTMs.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a6c81e69fc6a6113cc3a1c8af25ce7bffeea20dddaaab4e40d518b093ab378b7"},"source":{"id":"1609.09106","kind":"arxiv","version":4},"verdict":{"id":"ffe8543a-bf39-408a-baa3-3f8263dc554f","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T00:14:34.273271Z","strongest_claim":"Our main result is that hypernetworks can generate non-shared weights for LSTM and achieve near state-of-the-art results on a variety of sequence modelling tasks including character-level language modelling, handwriting generation and neural machine translation, challenging the weight-sharing paradigm for recurrent networks.","one_line_summary":"Hypernetworks generate weights for a main network, allowing LSTMs to use non-shared weights and achieve near state-of-the-art results on sequence modeling tasks while using fewer parameters overall.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the hypernetwork can be effectively trained end-to-end with backpropagation to produce high-quality weights for the main network without introducing instability, overfitting, or requiring excessive additional computation.","pith_extraction_headline":"A hypernetwork generates the weights for another network to enable non-shared weights in LSTMs."},"references":{"count":2,"sample":[{"doi":"","year":2016,"title":"TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems","work_id":"91f3c09e-dae6-48ca-80c0-463dd1b1f6e1","ref_index":1,"cited_arxiv_id":"1603.04467","is_internal_anchor":false},{"doi":"","year":2015,"title":"Large Embedding","work_id":"f71b6e48-8bbe-460c-ba5d-4ad1d91006fd","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":2,"snapshot_sha256":"03304a60339b17fe679342dc4ba51791cc91c58006c5498e1c8726e97ce099f4","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"da15d5c9170af51bc94b5e34cd417ef8aebdb0b15b9d7e7f0373e8ce9a100ec2"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}