User-Centric Toxicity Detection with Multi-Modal Spiking Neural Networks and Policy Optimization
Bibliographic record
Abstract
Toxic interactions in online platforms pose significant challenges to user safety, inclusivity, and trust. While recent approaches have explored multi-modal and deep learning solutions, most fail to adapt to evolving user expectations and contextual nuances. This paper introduces a user-centric framework for toxicity detection that fuses text, audio, and visual data through Convolutional Spiking Neural Networks (CSNNs), optimized using a policy-gradient reinforcement learning strategy.By integrating late fusion mechanisms and spiking neurons, the system models both temporal and semantic aspects of communication. Policy optimization, specifically via Proximal Policy Optimization (PPO), enables dynamic adaptation based on user feedback and behavioral patterns, reducing overgeneralization and increasing contextual sensitivity. Experiments on the ToxVidLM dataset demonstrate improved accuracy, robustness, and user-aligned detection performance compared to state-of-the-art models.This work offers a scalable, real-time architecture tailored for safety-sensitive environments such as live streaming, virtual meetings, and online gaming, marking a step toward intelligent, user-aware toxicity moderation tools.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".