Qwen3 causal-LM architecture
The captured configuration identifies Qwen3ForCausalLM with model type qwen3 and Transformers support.
Open Source Model Profile · MemTensor
MemOperator-4B
MemOperator-4B is a 4.02B-parameter Qwen3-family text-generation fine-tune from MemTensor. According to the model card, it is built for MemOS memory extraction and reorganization.
- Publisher
- MemTensor
- Task
- text-generation
- Model type
- qwen3
- License
- apache-2.0
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
MemOperator-4B is published by MemTensor as a text-generation model. The captured configuration identifies Qwen3ForCausalLM with model type qwen3, and Safetensors metadata reports about 4.02B parameters. According to the model card, it is the 4B MemOS memory-operator model fine-tuned from the Qwen3 series for memory extraction, integration, and update.
Recorded capabilities
4.02B MemOperator scale
Safetensors metadata reports 4,022,468,096 parameters; according to the model card, this is the 4B entry in the 4B, 1.7B, and 0.6B series fine-tuned from Qwen3.
MemOS memory-operation focus
According to the model card, the model handles memory extraction, integration, and update, with current support for extraction and clustering-based reorganization.
SFT with bilingual and context documentation
According to the model card, the series uses supervised fine-tuning on human-annotated and model-generated data, supports English and Chinese, and documents 32,768-token context.
Hugging Face, vLLM, and SGLang loading
According to the model card, the model can be loaded directly via Hugging Face, vLLM, or SGLang with preset extraction templates.
Use cases in the source record
- MemOS memory-extraction and clustering-based reorganization workflows using the publisher-documented configuration.
- Local-only MemOS deployment experiments consistent with the publisher-stated restricted-environment objective.
Limitations and unknowns
- No evaluation results were extracted in structured form.
- According to the model card, the publisher claims the 4B model surpasses GPT-4o-mini while remaining consumer-hardware deployable; this is a publisher claim, not an Ethen-verified result.
- According to the model card, conflict resolution and relational reasoning are work in progress.
- Provider state is historical snapshot data and should be refreshed before being presented as current.
Source and provenance
Source: MemTensor/MemOperator-4B
Captured: Unknown. Processed: 2026-09-07T19:34:34.278571+00:00.
Model Overview Memory Operator is a specialized language model developed for MemOS, designed to handle memory-related operations. Its core capabilities include memory extraction, integration, and update . The primary objectives for developing the Memory Operator sub-model are: Support local-only deployment , enabling the use of MemOS in restricted environments where internet connectivity is unavailable. Achieve memory operations at lower cost and higher speed , while maintaining high system performance. We are releasing the MemOperator model series in three sizes: 4B, 1.7B, and 0.6B parameters . These models are fine-tuned from the…
F001F002F003F004F005F006F007F008F010F011F013F014F015F016F017F018