← Back to Leaderboard

baidu logoBaidu | Ernie Light Dt 1.0 Intl

ERNIE-Light-DT-1.0 is based on open-source GLM models and includes research findings from Baidu Search's Lightning large model. It features Baidu Sparse Attention (BSA), a unique sparse attention mechanism developed by Baidu ERNIE. By using feedback-driven history compression, it improves the storage and indexing of historical information. This significantly reduces KV cache storage and computational costs while maintaining model performance. As a result, it allows for longer context windows at a lower cost.

Compare

Fair

47.7
Most Recent Test

Significant guardrail issues may impede work.

Strengths

  • Excellent at 2.2. Universality of Sin
  • Excellent at 3.6. Salvation Through Faith

Weaknesses

  • Limited task completion capability
  • Struggles with 1.7. Difficult Passages
  • Struggles with 2.1. Exclusivity of Jesus
  • Struggles with 3.1. Existence of God
Overall Score
47.7
Tier 1 (Task) 70%
43.8
Tier 2 (Doctrine) 20%
53.3
Tier 3 (Worldview) 10%
63.3
Performance Profile
Visual representation of performance across all evaluated categories
Model Information
Description

ERNIE-Light-DT-1.0 is based on open-source GLM models and includes research findings from Baidu Search's Lightning large model. It features Baidu Sparse Attention (BSA), a unique sparse attention mechanism developed by Baidu ERNIE. By using feedback-driven history compression, it improves the storage and indexing of historical information. This significantly reduces KV cache storage and computational costs while maintaining model performance. As a result, it allows for longer context windows at a lower cost.

Provider

Baidu

Model ID

baidu/ernie-light-dt-1.0-intl

Tests Run

1

Insights & Analysis

Categories

Category Heatmap
Performance breakdown by category - darker green indicates stronger alignment
43
1.1
40
1.2
47
1.3
43
1.4
47
1.5
50
1.6
37
1.7
20
2.1
80
2.2
60
2.3
60
2.4
50
2.5
50
2.6
25
3.1
75
3.2
63
3.3
75
3.4
50
3.5
83
3.6
Low
High
Category Breakdown (Bar Chart)
Performance across different categories
1.x = Task Capability2.x = Gospel Core3.x = Worldview Confession

Recent Tests

Recent Test Runs
DateScoreVersionTier 1 (Task)Tier 2 (Gospel)Tier 3 (Worldview)Trust Tier
11/11/202647.71.0.043.853.363.3automated