Radar ·
onPanda: model annotation and inspection
Published on September 22, 2026, this entry describes an open-source tool for annotating LLM data and inspecting models, including token-level edits and inspection of token probabilities and decoding.
On September 22, 2026, Radar featured onPanda, an open-source tool for annotating LLM data and inspecting models. The described workflow includes token-level edits, SFT and preference data, and inspection of token probabilities and decoding.
The post points to use by alignment engineers exploring on-policy data annotation and debugging model outputs. To check the scope and features, consult the original post and the project's public materials; verify which functions are available there and do not assume results beyond those described.
Original source: onPanda: anotação em nível de token e inspeção de modelos