Article
CommunityJev Models & Reimplementations
Jev-DLC: Decision-Level Continual Learning from Typed Feedback
arXiv paper on Decision-Level Continual Learning from Typed Feedback, a method that keeps Jev-style accept and revoke signals in bounded channels above a frozen scorer, and introduces TCPA-Bench, which credits a correction only through its effect on later decisions.
- Format
- Article
- Reading time
- 2 min
- Published by
- Community
- Added
- Oct 2026
Highlights
- Treats each typed correction as evidence about a recurring action identity plus a directed relation between the preferred and rejected choice.
- Retains no raw interaction history, adds no scorer call, and performs no runtime language-model backward pass.
- Introduces TCPA-Bench, a chronological suite that credits a correction only through its effect on later decisions.
- On TCPA-Repeat, reaches 44.83% user-macro accuracy overall, 2.43 points above Recent-4 history replay.
- After at least two earlier feedback events, reaches 56.07%, a 7.54-point margin over the strongest reported comparison at that depth.
Questions about this entry
Short answers drawn from the linked project or page.
It treats each correction as two signals: evidence about a recurring action identity, and a directed relation between the preferred and rejected choice. Those signals are preserved in complementary bounded channels above a frozen scorer and combined to shape the next decision.
No. The paper states the method retains no raw interaction history, adds no scorer call, and performs no runtime language-model backward pass, so corrections persist as compact state rather than as replayed transcripts.
TCPA-Bench is a chronological suite introduced alongside the method. It credits a correction only through its effect on later decisions, instead of measuring how well the correction itself is reproduced.
It reaches 44.83% user-macro accuracy overall on TCPA-Repeat, which is 2.43 points above Recent-4 history replay, and 56.07% after at least two earlier feedback events.
Once at least two earlier feedback events are present, Jev-DLC reaches 56.07% user-macro accuracy, a 7.54-point margin over the strongest reported comparison at that depth.
No. The method keeps typed feedback in bounded channels above a frozen scorer and combines their scores to shape decisions, with no runtime language-model backward pass and no additional scorer call.
The provided abstract describes the method and the TCPA-Bench evaluation suite with their reported numbers, but does not mention a released implementation, package, or install path.
More Jev resources like this
Open-weight vision-language judge that takes a task instruction plus 10 frames of a robot manipulation attempt and predicts success or failure with calibrated probabilities in a single forward pass. Released with inference code and a 20,628-prediction evaluation set.
Jev Models & ReimplementationsArticle#arxiv#open-models#evaluation
Community
arXiv paper proposing JevRec, which encodes recommendation items as variable-length semantic identifiers and adds a Jev-like decision modeling component with separate heads for length prediction and content generation, evaluated on three public datasets.
Jev Models & ReimplementationsArticle#arxiv#research#benchmark
Community
arXiv paper on a small candidate formalizer trained on offline teacher-generated lists that builds explicit, validated alternatives from task context for a frozen decision model, evaluated on CALVIN-D episodes and annotations.
Jev Models & ReimplementationsArticle#arxiv#research#benchmark
Community
A family of 0.8B-9B ophthalmic decision models that score candidate answers without autoregressive decoding, trained on RetiJev-Data: 78,158 clinical-decision and diagnosis questions and 50,074 guideline-grading questions. RetiJev-9B answers a case in 158 ms on one GPU.
Jev Models & ReimplementationsArticle#arxiv#open-models#benchmark
Community
arXiv paper introducing Flide, an exact inference engine for decision models such as Jev. It decouples the logical tree of a decision request from physical execution and reports answers up to 9.5 times faster than vLLM and SGLang.
Jev Models & ReimplementationsArticle#arxiv#benchmark#evaluation
Community
Apache-2.0 Hugging Face repository for RSI-JEV-27b, a 27B multimodal decision model based on Qwen3.8-27B that answers supplied multiple-choice questions over images, returning a selected option and probabilities for all options in one forward pass.
Jev Models & Reimplementations#open-models#benchmark#calibration
Community
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.