ramitpahwa13/voice-longmemeval
Voice-LongMemEval (v1) Benchmark for paralinguistic long-term memory: recalling how something was said (emotion, prosody, vocal events) across long multi-session conversations. A transcript alone is insufficient or misleading for every question. Extends LongMemEval (MIT; filler sessions reused verbatim). 316 questions (202 direct + 114 indirect) · 326 audio clips (76 min) · haystacks of 47 sessions (115k tokens) Browse it questions (default): one row per question… See the full description on the dataset page: https://huggingface.co/datasets/ramitpahwa13/voice-longmemeval.
| Author / organization | ramitpahwa13 |
|---|---|
| License (as declared) | cc-by-nc-sa-4.0 |
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 24 October 2026 |
| Last updated at source | 24 October 2026 |
Recent changes
- 24 Oct 2026 Description: Voice-LongMemEval (v1, 316 questions) Benchmark for paralinguistic long-term memory: recalling how something was said (emotion, prosody, vocal events) across long multi-session conversation histories. Extends LongMemEval (MIT; filler sessio → Voice-LongMemEval (v1) Benchmark for paralinguistic long-term memory: recalling how something was said (emotion, prosody, vocal events) across long multi-session conversations. A transcript alone is insufficient or misleading for every ques
- 24 Oct 2026 Topics added synthetic
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 24 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.