nielsr HF Staff commited on
Commit
e41e091
·
verified ·
1 Parent(s): aac9741

Add community evaluation results for Terminal-Bench 2.1, Terminal-Bench, SWE-bench Multilingual, HMMT February 2026

Browse files

This PR adds community-provided evaluation results for the following benchmarks:

- **[Terminal-Bench 2.1](https://huggingface.co/datasets/harborframework/terminal-bench-2.1?eval_result=thinkingmachines/Inkling-2-Small&leaderboard_task_id=terminalbench_2_1)**
- **[Terminal-Bench](https://huggingface.co/datasets/harborframework/terminal-bench?eval_result=thinkingmachines/Inkling-2-Small&leaderboard_task_id=terminalbench_4)**
- **[SWE-bench Multilingual](https://huggingface.co/datasets/SWE-bench/SWE-bench_Multilingual?eval_result=thinkingmachines/Inkling-2-Small&leaderboard_task_id=swe_bench_multilingual_%25_resolved)**
- **[HMMT February 2026](https://huggingface.co/datasets/MathArena/hmmt_feb_2026?eval_result=thinkingmachines/Inkling-2-Small&leaderboard_task_id=MathArena/hmmt_feb_2026)**

These results were extracted from the Inkling 2 Small model card on thinkingmachines.ai (Section 5 "Evaluations" table, Inkling 2 Small column, rows "Terminal Bench v2.1 (terminus v2)" 81.3%, "Terminal Bench v4.0 (mini-swe-agent v2.4.6)" 13.1%, "SWEBench Multilingual" 81.2%, "GPQA Diamond" 93.0% and "HMMT Feb 2026" 95.8%). These rows appear in the full table on thinkingmachines.ai but not in the shorter table in this repo's README. This is based on the new [evaluation results feature](https://huggingface.co/docs/hub/eval-results).

*Note: This is an automated PR. Please review the evaluation results before merging.*

.eval_results/hmmt_feb_2026.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: MathArena/hmmt_feb_2026
3
+ task_id: MathArena/hmmt_feb_2026
4
+ value: 95.8
5
+ source:
6
+ url: https://thinkingmachines.ai/model-card/inkling-2-small/
7
+ name: Inkling 2 Small model card
.eval_results/swe-bench_multilingual.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: SWE-bench/SWE-bench_Multilingual
3
+ task_id: swe_bench_multilingual_%_resolved
4
+ value: 81.2
5
+ source:
6
+ url: https://thinkingmachines.ai/model-card/inkling-2-small/
7
+ name: Inkling 2 Small model card
.eval_results/terminal-bench-2.1.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: harborframework/terminal-bench-2.1
3
+ task_id: terminalbench_2_1
4
+ value: 81.3
5
+ notes: "Terminus 2 harness"
6
+ source:
7
+ url: https://thinkingmachines.ai/model-card/inkling-2-small/
8
+ name: Inkling 2 Small model card
.eval_results/terminal-bench.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: harborframework/terminal-bench
3
+ task_id: terminalbench_4
4
+ value: 13.1
5
+ notes: "Terminal-Bench v4.0, mini-swe-agent v2.4.6 harness"
6
+ source:
7
+ url: https://thinkingmachines.ai/model-card/inkling-2-small/
8
+ name: Inkling 2 Small model card