Open-Source LLM Helps Safeguard Text Generation Prompts and Responses (opens in new tab)

(corp.roblox.com)

2 pointsmoneil9718mo ago1 comments

1 comments

Roblox’s LLM is currently outperforming popular LLM guardrail models on standard benchmarks. Roblox open-sourced both the LLM weights and the RoGuard-Eval benchmarking dataset.

j / k navigate · click thread line to collapse

Open-Source LLM Helps Safeguard Text Generation Prompts and Responses (opens in new tab)

(corp.roblox.com)

2 pointsmoneil9718mo ago1 comments

1 comments

moneil971OP8mo ago

Roblox’s LLM is currently outperforming popular LLM guardrail models on standard benchmarks. Roblox open-sourced both the LLM weights and the RoGuard-Eval benchmarking dataset.

j / k navigate · click thread line to collapse