aiqbench takeAI Products

Knowledgator's GLiFormer introduces a schema-conditioned encoder that performs nested JSON extraction without autoregressive token generation, significantly reducing inference latency compared to standard LLMs. By grounding all output values directly in source text spans, the model eliminates hallucinations while maintaining high F1 scores on complex extraction tasks. This architecture represents a meaningful shift toward efficient, deterministic information extraction pipelines rather than relying solely on generative capabilities.

Source coveringMarkTechPost

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Knowledgator Releases GLiFormer: A 575M-Parameter Encoder That Hits 91.10 F1 on Nested JSON Extraction Without Generating Tokens

MarkTechPostSeptember 16, 20264 min read

GLiFormer Large scores 91.10 F1 on nested JSON, near GPT-5.6-luna's 91.96, while grounding every value in source spans. The post Knowledgator Releases GLiFormer: A 575M-Parameter Encoder That Hits 91.10 F1 on Nested JSON Extraction Without Generating Tokens appeared first on MarkTechPost.

Knowledgator Engineering has released GLiFormer, a schema-conditioned encoder framework for information extraction. One model handles named-entity recognition (NER), text classification, relation extraction, nested JSON structuring, and text embeddings. You pass labels and extraction schemas at inference time. Two checkpoints are on Hugging Face. GLiFormer Base v1 has 264.2M parameters, and GLiFormer Large v1 has 575.6M. Deployable today? Yes. Both checkpoints are Apache 2.0, install with pip install gliformer, and run on CPU or…

Excerpt shown under the publisher's RSS terms · 73 words · full text and figures are on the publisher's site

Read at MarkTechPost ↗Ask about this article