# Long-context reasoning and retrieval on LoCoMo (test)

Source: [Long-context reasoning and retrieval on LoCoMo (test)](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:04:23.932Z

Evaluation of model performance on the test split of the LoCoMo dataset for long-context reasoning and retrieval tasks, measured using Overall, Single-Hop, Multi-Hop, Temporal, and Open Domain F1 and BLEU-1 scores.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Long-context reasoning and retrieval |
| Dataset (including any stated split) | LoCoMo (test) |
| Metric | Single-Hop F1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 37 of 37 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Single-Hop F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Single-Hop%20F1>): Higher is better; unit: PERCENT.

- [Overall F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Overall%20F1>): Higher is better; unit: PERCENT.

- [Overall BLEU-1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Overall%20BLEU-1>): Higher is better; unit: PERCENT.

- [Single-Hop BLEU-1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Single-Hop%20BLEU-1>): Higher is better; unit: PERCENT.

- [Multi-Hop F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Multi-Hop%20F1>): Higher is better; unit: PERCENT.

- [Multi-Hop BLEU-1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Multi-Hop%20BLEU-1>): Higher is better; unit: PERCENT.

- [Temporal F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Temporal%20F1>): Higher is better; unit: PERCENT.

- [Temporal BLEU-1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Temporal%20BLEU-1>): Higher is better; unit: PERCENT.

- [Open Domain F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Open%20Domain%20F1>): Higher is better; unit: PERCENT.

- [Open Domain BLEU-1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Open%20Domain%20BLEU-1>): Higher is better; unit: PERCENT.

- [Overall Accuracy](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Overall%20Accuracy>): Higher is better; unit: PERCENT.

- [Commonsense F1](<https://www.sota2.com/research/sota/long-context-reasoning-and-retrieval-on-locomo-test.md?metric=Commonsense%20F1>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Single-Hop F1 | Result date | Report paper |
| --- | --- | --- | --- |
| TraceMem | 95.12 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| TraceMem | 93.1 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| FC | 92.2 | Not recorded | 6965d1886343fa78c8323593 |
| FullText | 92.03 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| EM+SM | 89 | Not recorded | 6965d1886343fa78c8323593 |
| Nemori | 87.04 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| EM | 86.8 | Not recorded | 6965d1886343fa78c8323593 |
| ReadAgent | 85.7 | Not recorded | 6965d1886343fa78c8323593 |
| FullText | 83.95 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| Nemori | 82.52 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| LightMem | 81.21 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| LightMem | 74.14 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| RAG | 70 | Not recorded | 6965d1886343fa78c8323593 |
| Mem0 | 65.7 | Not recorded | 6965d1886343fa78c8323593 |
| NaiveRAG | 62.07 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| A-Mem | 61.47 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| E-mem | 61.14 | Not recorded | 697c3b5da8b01e76cc8da676 |
| E-mem | 59.23 | Not recorded | 697c3b5da8b01e76cc8da676 |
| NaiveRAG | 56.6 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| GAM | 56.35 | Not recorded | 697c3b5da8b01e76cc8da676 |
| HippoRAG | 55.7 | Not recorded | 6965d1886343fa78c8323593 |
| RAG | 52.45 | Not recorded | 697c3b5da8b01e76cc8da676 |
| MEMORYOS | 48.62 | Not recorded | 697c3b5da8b01e76cc8da676 |
| A-Mem | 47.92 | Not recorded | 698c0d5cc97a0b3e3b5e13de |
| RAG | 47.87 | Not recorded | 697c3b5da8b01e76cc8da676 |
| GAM | 47.74 | Not recorded | 697c3b5da8b01e76cc8da676 |
| Mem0 | 47.65 | Not recorded | 697c3b5da8b01e76cc8da676 |
| Long-Context | 46.68 | Not recorded | 697c3b5da8b01e76cc8da676 |
| MEMORYOS | 46.33 | Not recorded | 697c3b5da8b01e76cc8da676 |
| Long-Context | 46.05 | Not recorded | 697c3b5da8b01e76cc8da676 |
| A-mem | 44.65 | Not recorded | 697c3b5da8b01e76cc8da676 |
| A-MEM | 44.3 | Not recorded | 6965d1886343fa78c8323593 |
| Mem0 | 42.58 | Not recorded | 697c3b5da8b01e76cc8da676 |
| LIGHTMEM | 41.79 | Not recorded | 697c3b5da8b01e76cc8da676 |
| LIGHTMEM | 34.92 | Not recorded | 697c3b5da8b01e76cc8da676 |
| A-mem | 33.75 | Not recorded | 697c3b5da8b01e76cc8da676 |
| Zep | 32.9 | Not recorded | 6965d1886343fa78c8323593 |

### TraceMem

Single-Hop F1: 95.12. Result ID: 69963cfb5698665ec48009f3.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### TraceMem

Single-Hop F1: 93.1. Result ID: 69963cfb5698665ec48009ed.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### FC

Single-Hop F1: 92.2. Result ID: 6996926b5698665ec4db9135.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

Recorded evaluation configuration: {"latency p50":"4.18s","latency p90":"6.08s","latency p95":"6.92s","latency p99":"9.35s"}. This record does not assert independent reproduction.

### FullText

Single-Hop F1: 92.03. Result ID: 69963cfb5698665ec48009ee.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### EM+SM

Single-Hop F1: 89. Result ID: 6996926b5698665ec4db9137.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

Recorded evaluation configuration: {"latency p50":"2.28s","latency p90":"3.21s","latency p95":"3.68s","latency p99":"4.18s"}. This record does not assert independent reproduction.

### Nemori

Single-Hop F1: 87.04. Result ID: 69963cfb5698665ec48009f2.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7add1e209131cea85932.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### EM

Single-Hop F1: 86.8. Result ID: 6996926b5698665ec4db9136.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

Recorded evaluation configuration: {"latency p50":"2.19s","latency p90":"2.94s","latency p95":"3.23s","latency p99":"3.84s"}. This record does not assert independent reproduction.

### ReadAgent

Single-Hop F1: 85.7. Result ID: 6996926b5698665ec4db9134.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74351e209131cea6074c.md>)

Recorded evaluation configuration: {"latency p50":"10.59s","latency p90":"15.33s","latency p95":"18.00s","latency p99":"25.50s"}. This record does not assert independent reproduction.

### FullText

Single-Hop F1: 83.95. Result ID: 69963cfb5698665ec48009e8.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### Nemori

Single-Hop F1: 82.52. Result ID: 69963cfb5698665ec48009ec.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7add1e209131cea85932.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### LightMem

Single-Hop F1: 81.21. Result ID: 69963cfb5698665ec48009f1.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c6d1e209131cea8e04d.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### LightMem

Single-Hop F1: 74.14. Result ID: 69963cfb5698665ec48009eb.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c6d1e209131cea8e04d.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### RAG

Single-Hop F1: 70. Result ID: 6996926b5698665ec4db9132.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

Recorded evaluation configuration: {"latency p50":"1.50s","latency p90":"2.15s","latency p95":"2.40s","latency p99":"3.17s"}. This record does not assert independent reproduction.

### Mem0

Single-Hop F1: 65.7. Result ID: 6996926b5698665ec4db912f.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa784b1e209131cea77ad6.md>)

Recorded evaluation configuration: {"latency p50":"1.04s","latency p90":"1.63s","latency p95":"1.81s","latency p99":"2.07s"}. This record does not assert independent reproduction.

### NaiveRAG

Single-Hop F1: 62.07. Result ID: 69963cfb5698665ec48009ef.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### A-Mem

Single-Hop F1: 61.47. Result ID: 69963cfb5698665ec48009f0.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf31e209131cea8b869.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4.1-mini"}. This record does not assert independent reproduction.

### E-mem

Single-Hop F1: 61.14. Result ID: 6996184c5698665ec4581b1d.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### E-mem

Single-Hop F1: 59.23. Result ID: 6996184c5698665ec4581b15.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### NaiveRAG

Single-Hop F1: 56.6. Result ID: 69963cfb5698665ec48009e9.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### GAM

Single-Hop F1: 56.35. Result ID: 6996184c5698665ec4581b1c.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6925380fd6b0dd334292f41a.md>)

[Implementation](<https://github.com/VectorSpaceLab/general-agentic-memory>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### HippoRAG

Single-Hop F1: 55.7. Result ID: 6996926b5698665ec4db9133.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76981e209131cea6dff9.md>)

Recorded evaluation configuration: {"latency p50":"3.51s","latency p90":"6.01s","latency p95":"7.00s","latency p99":"9.72s"}. This record does not assert independent reproduction.

### RAG

Single-Hop F1: 52.45. Result ID: 6996184c5698665ec4581b0f.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### MEMORYOS

Single-Hop F1: 48.62. Result ID: 6996184c5698665ec4581b12.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793f1e209131cea7cdc0.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### A-Mem

Single-Hop F1: 47.92. Result ID: 69963cfb5698665ec48009ea.

[Reporting paper](<https://www.sota2.com/research/paper/698c0d5cc97a0b3e3b5e13de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf31e209131cea8b869.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### RAG

Single-Hop F1: 47.87. Result ID: 6996184c5698665ec4581b17.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### GAM

Single-Hop F1: 47.74. Result ID: 6996184c5698665ec4581b14.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6925380fd6b0dd334292f41a.md>)

[Implementation](<https://github.com/VectorSpaceLab/general-agentic-memory>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### Mem0

Single-Hop F1: 47.65. Result ID: 6996184c5698665ec4581b11.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa784b1e209131cea77ad6.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### Long-Context

Single-Hop F1: 46.68. Result ID: 6996184c5698665ec4581b0e.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### MEMORYOS

Single-Hop F1: 46.33. Result ID: 6996184c5698665ec4581b1a.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa793f1e209131cea7cdc0.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### Long-Context

Single-Hop F1: 46.05. Result ID: 6996184c5698665ec4581b16.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### A-mem

Single-Hop F1: 44.65. Result ID: 6996184c5698665ec4581b10.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf31e209131cea8b869.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### A-MEM

Single-Hop F1: 44.3. Result ID: 6996926b5698665ec4db9130.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf31e209131cea8b869.md>)

Recorded evaluation configuration: {"latency p50":"0.77s","latency p90":"1.06s","latency p95":"1.25s","latency p99":"2.09s"}. This record does not assert independent reproduction.

### Mem0

Single-Hop F1: 42.58. Result ID: 6996184c5698665ec4581b19.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa784b1e209131cea77ad6.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### LIGHTMEM

Single-Hop F1: 41.79. Result ID: 6996184c5698665ec4581b13.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c6d1e209131cea8e04d.md>)

Recorded evaluation configuration: {"Backbone":"GPT-4o-mini"}. This record does not assert independent reproduction.

### LIGHTMEM

Single-Hop F1: 34.92. Result ID: 6996184c5698665ec4581b1b.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c6d1e209131cea8e04d.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### A-mem

Single-Hop F1: 33.75. Result ID: 6996184c5698665ec4581b18.

[Reporting paper](<https://www.sota2.com/research/paper/697c3b5da8b01e76cc8da676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf31e209131cea8b869.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-14B"}. This record does not assert independent reproduction.

### Zep

Single-Hop F1: 32.9. Result ID: 6996926b5698665ec4db9131.

[Reporting paper](<https://www.sota2.com/research/paper/6965d1886343fa78c8323593.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e944.md>)

Recorded evaluation configuration: {"latency p50":"1.25s","latency p90":"1.65s","latency p95":"1.74s","latency p99":"2.27s"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
