Unlocking Cognitive Capabilities and Analyzing the Perception-Logic Trade-off

TL;DR

This paper introduces MERaLiON2-Omni, a multilingual multimodal model designed for Southeast Asia, which explicitly separates perception and reasoning modules, and analyzes the trade-offs and challenges in integrating sensory grounding with complex reasoning.

Contribution

It presents a novel training pipeline that decouples perception and reasoning, introduces a region-specific multimodal dataset, and provides a detailed analysis of the perception-logic trade-off in multimodal models.

Findings

Reasoning enhances performance on abstract tasks but causes instability in sensory processing.

Identifies temporal drift in audio and visual over-interpretation as key issues.

Proposes a cost-effective data synthesis pipeline using a Super-LLM.

Abstract

Recent advancements in Multimodal Large Language Models (MLLMs) pursue omni-perception capabilities, yet integrating robust sensory grounding with complex reasoning remains a challenge, particularly for underrepresented regions. In this report, we introduce the research preview of MERaLiON2-Omni (Alpha), a 10B-parameter multilingual omni-perception tailored for Southeast Asia (SEA). We present a progressive training pipeline that explicitly decouples and then integrates "System 1" (Perception) and "System 2" (Reasoning) capabilities. First, we establish a robust Perception Backbone by aligning region-specific audio-visual cues (e.g., Singlish code-switching, local cultural landmarks) with a multilingual LLM through orthogonal modality adaptation. Second, to inject cognitive capabilities without large-scale supervision, we propose a cost-effective Generate-Judge-Refine pipeline. By utilizing a Super-LLM to filter hallucinations and resolve conflicts via a consensus mechanism, we synthesize high-quality silver data that transfers textual Chain-of-Thought reasoning to multimodal scenarios. Comprehensive evaluation on our newly introduced SEA-Omni Benchmark Suite reveals an Efficiency-Stability Paradox: while reasoning acts as a non-linear amplifier for abstract tasks (boosting mathematical and instruction-following performance significantly), it introduces instability in low-level sensory processing. Specifically, we identify Temporal Drift in long-context audio, where extended reasoning desynchronizes the model from acoustic timestamps, and Visual Over-interpretation, where logic overrides pixel-level reality. This report details the architecture, the data-efficient training recipe, and a diagnostic analysis of the trade-offs between robust perception and structured reasoning.