Spaces:

ParetoOptimal
/

repro-plugmem

Running

14.6 kB

Ctrl+K

1 contributor

Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached).

bf5cbdc verified 3 days ago

00-paper-and-approach
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
01-executive-summary
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
02-claim-1-dense-tokens
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
03-claim-2-units
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
04-claim-3-capabilities
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
05-claim-4-longmemeval
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
99-conclusion
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago
index.md

876 Bytes
Rebuild NWKaQIKoGp (PlugMem): GENUINE reproduction on the real LongMemEval benchmark (xiaowu0162/longmemeval, N=60) with Qwen2.5-7B via free HF router. PlugMem beats raw-history baseline at equal budget (26.7% vs 21.7%) at fewer tokens (310 vs 357); 852 semantic + 118 procedural units; task-agnostic across 6 question types. $0, deterministic (cached). 3 days ago