Datasets:
agent stringclasses 1
value | benchmark stringclasses 1
value | context dict | context_length_error bool 1
class | exception dict | finished_at stringlengths 27 27 | harness_fallback_turn_count int64 0 22 | malformed_assistant_turn_count int64 0 0 | max_completion_tokens_per_turn int64 0 15k | max_prompt_tokens_per_turn int64 0 32.7k | message_schema_version stringclasses 1
value | messages listlengths 1 149 | messages_without_thinking listlengths 1 149 | model stringclasses 1
value | model_source stringclasses 1
value | n_agent_turns int64 0 74 | n_cache_tokens int64 0 0 | n_input_tokens int64 0 544k | n_output_tokens int64 0 55.2k | n_test_cases_passed int64 0 10 | n_test_cases_total int64 0 11 | official_error_zero bool 2
classes | output_length_cap_events int64 0 0 | raw_reward float64 0 1 ⌀ | reward float64 0 1 | sampling dict | source_run stringclasses 1
value | started_at stringlengths 27 27 | steps listlengths 1 78 | steps_without_thinking listlengths 1 78 | summarization_count int64 0 44 | task stringclasses 89
values | test_cases listlengths 0 11 | thinking_markup_complete bool 2
classes | thinking_markup_policy stringclasses 1
value | thinking_open_tag_inferred bool 2
classes | thinking_turn_count int64 0 71 | trajectory_available bool 1
class | trajectory_integrity dict | trial_id stringlengths 16 41 | untagged_assistant_turn_count int64 0 18 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 7200.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T12:01:47.968749Z | 11 | 0 | 951 | 32,237 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 41 | 0 | 191,518 | 7,792 | 0 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T10:01:03.133769Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 24 | bn-fit-modify | [{"name":"test_outputs.py::test_bn_sample_exists","status":"failed"},{"name":"test_outputs.py::test_(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 41 | true | {"base_present":true,"contiguous":false,"continuation_indices":[24],"missing_continuation_indices":[(...TRUNCATED) | bn-fit-modify__psppHmG | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | null | 2026-08-01T07:52:04.335893Z | 22 | 0 | 4,575 | 30,849 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 38 | 0 | 89,777 | 10,087 | 0 | 3 | false | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T07:29:59.909579Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 44 | make-mips-interpreter | [{"name":"test_outputs.py::test_vm_execution","status":"failed"},{"name":"test_outputs.py::test_fram(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 38 | true | {"base_present":true,"contiguous":false,"continuation_indices":[44],"missing_continuation_indices":[(...TRUNCATED) | make-mips-interpreter__6yhcUq4 | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | null | 2026-08-01T02:17:27.159149Z | 12 | 0 | 5,500 | 32,490 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 68 | 0 | 276,663 | 15,547 | 0 | 3 | false | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T01:53:12.644424Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 23 | make-mips-interpreter | [{"name":"test_outputs.py::test_vm_execution","status":"failed"},{"name":"test_outputs.py::test_fram(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 68 | true | {"base_present":true,"contiguous":false,"continuation_indices":[3,23],"missing_continuation_indices"(...TRUNCATED) | make-mips-interpreter__P8vKXSk | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | null | 2026-08-01T08:08:46.945422Z | 14 | 0 | 2,456 | 32,703 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 74 | 0 | 357,279 | 16,739 | 2 | 5 | false | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T07:30:00.015288Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 30 | video-processing | [{"name":"test_outputs.py::test_example_video_exists","status":"passed"},{"name":"test_outputs.py::t(...TRUNCATED) | false | preserve-zero-tag-generations-without-inference | true | 63 | true | {"base_present":true,"contiguous":false,"continuation_indices":[1,2,30],"missing_continuation_indice(...TRUNCATED) | video-processing__fSfJ9BV | 11 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | null | 2026-08-01T09:21:31.236929Z | 15 | 0 | 5,309 | 32,712 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 71 | 0 | 301,117 | 17,681 | 3 | 5 | false | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T07:29:59.607117Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 32 | video-processing | [{"name":"test_outputs.py::test_example_video_exists","status":"passed"},{"name":"test_outputs.py::t(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 71 | true | {"base_present":true,"contiguous":false,"continuation_indices":[1,2,32],"missing_continuation_indice(...TRUNCATED) | video-processing__my99Bjx | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 1800.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T08:01:13.590210Z | 0 | 0 | 3,229 | 16,952 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 10 | 0 | 106,678 | 10,646 | 7 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T07:29:59.272930Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 0 | adaptive-rejection-sampler | [{"name":"test_outputs.py::test_ars_function_exists","status":"passed"},{"name":"test_outputs.py::te(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 10 | true | {
"base_present": true,
"contiguous": true,
"continuation_indices": [],
"missing_continuation_indices": []
} | adaptive-rejection-sampler__gN7hQBg | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 1800.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T08:31:45.133463Z | 0 | 0 | 6,519 | 20,514 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 25 | 0 | 288,701 | 19,251 | 7 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T08:00:58.093712Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 1 | adaptive-rejection-sampler | [{"name":"test_outputs.py::test_ars_function_exists","status":"passed"},{"name":"test_outputs.py::te(...TRUNCATED) | false | preserve-zero-tag-generations-without-inference | true | 24 | true | {
"base_present": true,
"contiguous": true,
"continuation_indices": [
1
],
"missing_continuation_indices": []
} | adaptive-rejection-sampler__iegvCJD | 1 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 1800.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T01:50:41.440059Z | 0 | 0 | 3,837 | 27,836 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 65 | 0 | 341,171 | 38,080 | 3 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T01:19:28.668178Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 3 | adaptive-rejection-sampler | [{"name":"test_outputs.py::test_ars_function_exists","status":"passed"},{"name":"test_outputs.py::te(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 65 | true | {"base_present":true,"contiguous":true,"continuation_indices":[1,2,3],"missing_continuation_indices"(...TRUNCATED) | adaptive-rejection-sampler__surCvgY | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 1800.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T08:06:35.784431Z | 0 | 0 | 6,034 | 22,766 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 18 | 0 | 142,841 | 22,033 | 4 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T07:33:57.971739Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 1 | adaptive-rejection-sampler | [{"name":"test_outputs.py::test_ars_function_exists","status":"passed"},{"name":"test_outputs.py::te(...TRUNCATED) | true | preserve-zero-tag-generations-without-inference | true | 18 | true | {
"base_present": true,
"contiguous": true,
"continuation_indices": [
1
],
"missing_continuation_indices": []
} | adaptive-rejection-sampler__u8pc2nP | 0 |
terminus-2 | terminal-bench/terminal-bench-2-1 | {"interleaved_thinking":false,"model_context_length":65536,"proactive_summarization_threshold":40960(...TRUNCATED) | false | {
"message": "Agent execution timed out after 7200.0 seconds",
"type": "AgentTimeoutError"
} | 2026-08-01T12:01:44.868875Z | 0 | 0 | 15,016 | 26,174 | 1.1 | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"content":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | qwen35-action-only-20k | violetxi/qwen35-4b-offline-echo-action-only-20k-tacc | 37 | 0 | 502,026 | 55,239 | 0 | 9 | true | 0 | 0 | 0 | {"enable_thinking":true,"extra_body":{"chat_template_kwargs":{"enable_thinking":true},"min_p":0.0,"r(...TRUNCATED) | qwen35-action-only-20k_infra-repaired-c164-20260801T072925Z | 2026-08-01T10:01:02.715631Z | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | [{"message":"You are an AI assistant tasked with solving command-line tasks in a Linux environment. (...TRUNCATED) | 3 | bn-fit-modify | [{"name":"test_outputs.py::test_bn_sample_exists","status":"failed"},{"name":"test_outputs.py::test_(...TRUNCATED) | false | preserve-zero-tag-generations-without-inference | true | 28 | true | {"base_present":true,"contiguous":true,"continuation_indices":[1,2,3],"missing_continuation_indices"(...TRUNCATED) | bn-fit-modify__BRpXACR | 9 |
End of preview. Expand in Data Studio
qwen35-action-only-20k — Terminal-Bench 2.1
Noncanonical Terminal-Bench 2.1 evaluation of violetxi/qwen35-4b-offline-echo-action-only-20k-tacc through the served
model ID qwen35-action-only-20k with Terminus-2.
Noncanonical run: timeout_multiplier=2 instead of 1.0; repair concurrency=164 exceeds 30. Do not compare this score directly with canonical TB2.1 leaderboard runs.
Result
- Recorded trials: 445
- Tasks / attempts: 89 × 5
- Errored trials scored as zero: 250
- Exception counts:
{"AgentTimeoutError": 249, "InternalServerError": 0, "VerifierTimeoutError": 1} - Agent timeouts / context-length events / output-cap events: 249 / 0 / 0
- Mean reward / Pass@1: 0.119101
- Pass@5: 0.258427
- Total input/output/cache tokens: 70960809 / 3875752 / 0
- Proactive summaries: 273
Protocol
Temperature 0.6, top-p
0.95, top-k omitted (disabled),
max_tokens=32768,
enable_thinking=true, max turns
30, concurrency 30 for the 78 preserved base trials and 164 for the 367 one-for-one replacement trials.
- Trials / untagged / malformed assistant turns with incomplete thinking markup: 76 / 267 / 0
- Assistant-turn thinking-markup completeness: 0.963782
The server context length was 65536;
Terminus proactive summarization began at
40960 tokens.
Task-native TB2.1 timeouts were scaled (timeout_multiplier=2.0), so this is a noncanonical evaluation. All errored trials remain in the denominator with reward
zero.
Infrastructure-repair provenance
- Preserved valid trials from the original run: 78
- One-for-one replacements for proven infrastructure-invalid trials: 367
- Composition policy: preserve valid trials; replace only proven infrastructure-invalid trials
- Scoring policy: errored trials count as reward 0
The replacement mapping, source paths, source run-config checksums, and
configuration-parity evidence are retained in run_config.json under
canonical_composition.
Viewer columns
messages and messages_without_thinking are aligned, strictly alternating
user/assistant conversations. Terminal observations are user messages.
steps and steps_without_thinking retain the detailed Harbor trajectory.
For thinking mode, Qwen's prompt-supplied opening <think> marker is
reconstructed while generated reasoning and </think> are preserved.
Zero-tag generations are preserved byte-for-byte in both views without
inference and are explicitly marked by thinking_markup_complete,
untagged_assistant_turn_count, and thinking_markup_policy.
trajectory_integrity reports which raw continuation segments were available;
missing segments are never fabricated.
Harbor's literal technical-difficulties fallback is marked with an empty
<think></think> block only in the normalized full-thinking view; its count is
recorded separately and the raw fallback text is retained.
Only verifier test names/statuses are published. Hidden verifier traces and source, authorization headers, API keys, and credential values are excluded or redacted.
- Downloads last month
- 49