WhitzardNUWA Lab ↗
NUWANUWA INDEX

Search NUWA Index

Research · Benchmarks · Findings · Perspectives

中文
← NUWA Index

Evidence record

Kimi K3

CharXiv · Evidence record

Sources and provenance ↓ Read the research narrative →

Source snapshot, chart and review scope →

Original source charts and materials

Recorded measurements

Verbalized evaluation awareness
36.5%

Per-result sample size not disclosed

Harness not specified · Budget not specified

Assessment provenance

Evaluating Kimi K3 for Misalignment Risks → · External report

Actual evaluation: Neo Research · Model record →

Method and scope

Tests understanding and reasoning over charts. Neo’s withheld-image adaptation measures attempts without the required input.

Evidence type
Benchmark
Assessment subject
The model or agent in the stated task configuration.
Primary dimension
Capability
Evidence snapshot
2026-10-29.1

The main dimension follows the stated measurement; contextual risk relevance does not establish harmful deployment capability.

Source finding

Verbalized evaluation awareness: 36.5 %. Conditions are retained from the source.

Limitations

The original visual capability benchmark and the missing-image behavioral test measure different constructs.

Only full reasoning-trace models analyzed; hidden/unexpressed awareness is not measured.

Related research questions

References

  1. [1] CharXiv
  2. [2] Evaluating Kimi K3 for Misalignment Risks. 2026-09-11. September 2026 ; §1.1.3 Table 1 p.11