BabyVision: Visual Reasoning Beyond Language
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 reserved pith:MZT5BHUOrecord.jsonopen to challenge →
read the original abstract
While humans develop core visual skills long before acquiring language, contemporary Multimodal LLMs (MLLMs) still rely heavily on linguistic priors to compensate for their fragile visual understanding. We uncovered a crucial fact: state-of-the-art MLLMs consistently fail on basic visual tasks that humans, even 3-year-olds, can solve effortlessly. To systematically investigate this gap, we introduce BabyVision, a benchmark designed to assess core visual abilities independent of linguistic knowledge for MLLMs. BabyVision spans a wide range of tasks, with 388 items divided into 22 subclasses across four key categories. Empirical results and human evaluation reveal that leading MLLMs perform significantly below human baselines. Gemini3-Pro-Preview scores 49.7, lagging behind 6-year-old humans and falling well behind the average adult score of 94.1. These results show despite excelling in knowledge-heavy evaluations, current MLLMs still lack fundamental visual primitives. Progress in BabyVision represents a step toward human-level visual perception and reasoning capabilities. We also explore solving visual reasoning with generation models by proposing BabyVision-Gen and automatic evaluation toolkit. Our code and benchmark data are released at https://github.com/UniPat-AI/BabyVision for reproduction.
This paper has not been read by Pith yet.
Forward citations
Cited by 21 Pith papers
-
Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients
ZPPO improves distillation to small vision-language models by using binary and negative candidate prompts plus a replay buffer for hard questions, outperforming standard distillation and GRPO on a 31-benchmark suite w...
-
LEVANTE-bench: Multi-Scale Comparison of VLMs to Children Using Cognitive Tasks (or, "Is Your VLM Smarter Than a 5th Grader?")
VLMs show partial alignment with children's performance on six cognitive tasks, with stronger models matching better at task and item levels but struggling on matrix reasoning and mental rotation.
-
ATLAS: Agentic Test-time Learning-to-Allocate Scaling
ATLAS introduces an LLM-orchestrated agentic framework for dynamic test-time scaling via extensible 'explore' actions, achieving higher accuracy with fewer API calls than fixed-workflow baselines on four benchmarks.
-
DeepLatent: Think with Images via Parallel Latent Visual Reasoning
DeepLatent introduces a parallel latent visual reasoning framework with learnable 2D tokens and continuous RL, trained via distillation then RL, plus a new 180K dataset, claiming SOTA benchmark results.
-
The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space
MLLMs scoring 70-83% on Cartesian visual tasks drop to 31-39% on logically equivalent polar versions, exposing reliance on grid discretization shortcuts instead of topology-invariant reasoning.
-
Bridging Structure and Language: Graph-Based Visual Reasoning for Autonomous Road Understanding
A graph-grounded Combined Road Substrate framework generates traceable QA pairs from road maps to improve small VLMs on compositional road reasoning tasks.
-
EgoBabyVLM: Benchmarking Cross-Modal Learning from Naturalistic Egocentric Video Data
Current VLMs depend on tightly aligned curated data and cannot exploit the weakly-aligned egocentric video signals that dominate naturalistic infant input.
-
Step-wise Rubric Rewards for LLM Reasoning
SRaR attributes rubric items to specific steps via an LLM judge, normalizes per-step scores across rollouts, and combines them with outcome rewards via a decoupled advantage estimator, yielding 3.57-point accuracy gai...
-
The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space
Reformulating 53 visual reasoning tasks in polar coordinates causes frontier MLLMs to drop from 70-83% to 31-39% accuracy while preserving logical equivalence, revealing a Cartesian shortcut in current benchmarks.
-
Do multimodal models imagine electric sheep?
Fine-tuning VLMs to output action sequences for puzzles causes emergent internal visual representations that improve performance when integrated into reasoning.
-
Injecting Distributional Awareness into MLLMs via Reinforcement Learning for Deep Imbalanced Regression
A plug-and-play RL method adds batch-level distributional supervision via CCC rewards to reduce regression-to-the-mean in MLLMs on imbalanced regression benchmarks.
-
Injecting Distributional Awareness into MLLMs via Reinforcement Learning for Deep Imbalanced Regression
A Group Relative Policy Optimization framework with concordance correlation coefficient rewards improves MLLM regression accuracy on long-tailed distributions, especially in medium- and few-shot regimes, without model...
-
The Expense of Seeing: Attaining Trustworthy Multimodal Reasoning Within the Monolithic Paradigm
Proposes the Modality Translation Protocol with metrics ToS, CoS, FoS and SSC to quantify visual knowledge bottlenecks in VLMs, plus a Divergence Law hypothesis that scaling language models may increase the penalty.
-
VLMs Need Words: Vision Language Models Ignore Visual Detail In Favor of Semantic Anchors
VLMs bypass visual comparison by recovering semantic labels for nameable entities and hallucinate on unnamable ones, as shown by performance gaps and Logit Lens analysis.
-
V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning
V-Zero trains MLLMs for visual reasoning without answer labels by gating on-policy distillation trajectories using contrastive evidence from relevant versus negative image crops.
-
VLMs Trace Without Tracking: Diagnosing Failures in Visual Path Following
VLMs frequently switch away from a target visual path to nearby similar distractors in controlled tracing tasks, with standard scaling, reasoning, and instruction interventions providing only partial mitigation.
-
SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture
SenseNova-U1 presents native unified multimodal models that match top understanding VLMs while delivering strong performance in image generation, infographics, and interleaved tasks via the NEO-unify architecture.
-
The Expense of Seeing: Attaining Trustworthy Multimodal Reasoning Within the Monolithic Paradigm
Vision-language models exhibit functional blindness by exploiting language priors over visual representations; the Modality Translation Protocol and metrics like Toll, Curse, and Fallacy of Seeing reveal this, support...
-
Kimi K2.5: Visual Agentic Intelligence
Kimi K2.5 combines joint text-vision training with an Agent Swarm parallel orchestration framework to reach claimed state-of-the-art results on coding, vision, reasoning, and agent tasks while cutting latency up to 4.5 times.
-
Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity
Seed2.0 model series reports gains in reasoning, visual understanding, search, and reliability on intricate long-horizon tasks via an internal evaluation system.
-
EXAONE 4.5 Technical Report
EXAONE 4.5 is a new open-weight multimodal model that matches general benchmarks and outperforms similar-scale models on document understanding and Korean contextual reasoning.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.