run directory (provenance + evals)

run/capability/capability_summary.csv

@@ -0,0 +1,3 @@
+benchmark,metric,m0,m1,delta_pp
+mmlu,"acc,none",67.72,68.42,+0.70
+gsm8k_cot,"exact_match,flexible-extract",61.00,68.00,+7.00

run/capability/capability_summary.md

@@ -0,0 +1,8 @@
+# Capability retention: M0 (base) vs M1 (resist SFT)
+
+arXiv:2511.21399 App. E.4 methodology — lm-eval-harness, MMLU 5-shot MC, GSM8K 8-shot CoT, greedy, accuracy on the test split.
+
+| benchmark | metric | M0 | M1 | delta (pp) |
+|---|---|---|---|---|
+| mmlu | acc,none | 67.7% | 68.4% | +0.7 |
+| gsm8k_cot | exact_match,flexible-extract | 61.0% | 68.0% | +7.0 |

run/capability/m0/gsm8k_cot/Qwen__Qwen2.5-3B-Instruct/results_2026-07-23T18-36-29.229735.json

@@ -0,0 +1,205 @@
+{
+  "results": {
+    "gsm8k_cot": {
+      "name": "gsm8k_cot",
+      "alias": "gsm8k_cot",
+      "sample_len": 200,
+      "exact_match,strict-match": 0.035,
+      "exact_match_stderr,strict-match": 0.013027801736688037,
+      "exact_match,flexible-extract": 0.61,
+      "exact_match_stderr,flexible-extract": 0.03457567623250012
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "gsm8k_cot": {
+      "task": "gsm8k_cot",
+      "dataset_path": "openai/gsm8k",
+      "dataset_name": "main",
+      "test_split": "test",
+      "doc_to_text": "Q: {{question}}\nA:",
+      "doc_to_target": "{{answer.split('####')[-1].strip() if answer is defined else target}}",
+      "unsafe_code": false,
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": [
+          {
+            "question": "There are 15 trees in the grove. Grove workers will plant trees in the grove today. After they are done, there will be 21 trees. How many trees did the grove workers plant today?",
+            "target": "There are 15 trees originally. Then there were 21 trees after some more were planted. So there must have been 21 - 15 = 6. The answer is 6."
+          },
+          {
+            "question": "If there are 3 cars in the parking lot and 2 more cars arrive, how many cars are in the parking lot?",
+            "target": "There are originally 3 cars. 2 more cars arrive. 3 + 2 = 5. The answer is 5."
+          },
+          {
+            "question": "Leah had 32 chocolates and her sister had 42. If they ate 35, how many pieces do they have left in total?",
+            "target": "Originally, Leah had 32 chocolates. Her sister had 42. So in total they had 32 + 42 = 74. After eating 35, they had 74 - 35 = 39. The answer is 39."
+          },
+          {
+            "question": "Jason had 20 lollipops. He gave Denny some lollipops. Now Jason has 12 lollipops. How many lollipops did Jason give to Denny?",
+            "target": "Jason started with 20 lollipops. Then he had 12 after giving some to Denny. So he gave Denny 20 - 12 = 8. The answer is 8."
+          },
+          {
+            "question": "Shawn has five toys. For Christmas, he got two toys each from his mom and dad. How many toys does he have now?",
+            "target": "Shawn started with 5 toys. If he got 2 toys each from his mom and dad, then that is 4 more toys. 5 + 4 = 9. The answer is 9."
+          },
+          {
+            "question": "There were nine computers in the server room. Five more computers were installed each day, from monday to thursday. How many computers are now in the server room?",
+            "target": "There were originally 9 computers. For each of 4 days, 5 more computers were added. So 5 * 4 = 20 computers were added. 9 + 20 is 29. The answer is 29."
+          },
+          {
+            "question": "Michael had 58 golf balls. On tuesday, he lost 23 golf balls. On wednesday, he lost 2 more. How many golf balls did he have at the end of wednesday?",
+            "target": "Michael started with 58 golf balls. After losing 23 on tuesday, he had 58 - 23 = 35. After losing 2 more, he had 35 - 2 = 33 golf balls. The answer is 33."
+          },
+          {
+            "question": "Olivia has $23. She bought five bagels for $3 each. How much money does she have left?",
+            "target": "Olivia had 23 dollars. 5 bagels for 3 dollars each will be 5 x 3 = 15 dollars. So she has 23 - 15 dollars left. 23 - 15 is 8. The answer is 8."
+          }
+        ],
+        "doc_to_text": "Q: {{question}}\nA:",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer.split('####')[-1].strip() if answer is defined else target}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 8,
+      "metric_list": [
+        {
+          "aggregation": "mean",
+          "higher_is_better": true,
+          "ignore_case": true,
+          "ignore_punctuation": false,
+          "metric": "exact_match",
+          "regexes_to_ignore": [
+            ",",
+            "\\$",
+            "(?s).*#### ",
+            "\\.$"
+          ]
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "do_sample": false,
+        "until": [
+          "Q:",
+          "</s>",
+          "<|im_end|>"
+        ]
+      },
+      "repeats": 1,
+      "filter_list": [
+        {
+          "filter": [
+            {
+              "function": "regex",
+              "regex_pattern": "The answer is (\\-?[0-9\\.\\,]+)."
+            },
+            {
+              "function": "take_first"
+            }
+          ],
+          "name": "strict-match"
+        },
+        {
+          "filter": [
+            {
+              "function": "regex",
+              "group_select": -1,
+              "regex_pattern": "(-?[$0-9.,]{2,})|(-?[0-9]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ],
+          "name": "flexible-extract"
+        }
+      ],
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "Qwen/Qwen2.5-3B-Instruct",
+        "dtype": "bfloat16",
+        "config_source": "/usr/local/lib/python3.11/dist-packages/lm_eval/tasks/gsm8k/gsm8k-cot.yaml"
+      }
+    }
+  },
+  "versions": {
+    "gsm8k_cot": 3.0
+  },
+  "n-shot": {
+    "gsm8k_cot": 8
+  },
+  "higher_is_better": {
+    "gsm8k_cot": {
+      "exact_match": true
+    }
+  },
+  "n-samples": {
+    "gsm8k_cot": {
+      "original": 1319,
+      "effective": 200
+    }
+  },
+  "config": {
+    "model": "hf",
+    "model_args": {
+      "pretrained": "Qwen/Qwen2.5-3B-Instruct",
+      "dtype": "bfloat16"
+    },
+    "model_num_parameters": 3085938688,
+    "model_dtype": "torch.bfloat16",
+    "model_revision": "main",
+    "model_sha": "aa8e72537993ba99e69dfaafa59ed015b17504d1",
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": 200.0,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {
+      "do_sample": false
+    },
+    "random_seed": 0,
+    "numpy_seed": 0,
+    "torch_seed": 0,
+    "fewshot_seed": 0
+  },
+  "git_hash": "eb4f2be22f7baf6d268c3dd5e46d49d6bd2e74ae",
+  "date": 1784830566.9120035,
+  "pretty_env_info": "PyTorch version: 2.6.0+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: Could not collect\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.11.10 (main, Sep  7 2024, 18:35:41) [GCC 11.4.0] (64-bit runtime)\nPython platform: Linux-6.8.0-52-generic-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: GPU 0: NVIDIA RTX A4000\nNvidia driver version: 550.144.03\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        48 bits physical, 48 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               112\nOn-line CPU(s) list:                  0-111\nVendor ID:                            AuthenticAMD\nModel name:                           AMD EPYC 7453 28-Core Processor\nCPU family:                           25\nModel:                                1\nThread(s) per core:                   2\nCore(s) per socket:                   28\nSocket(s):                            2\nStepping:                             1\nFrequency boost:                      enabled\nCPU max MHz:                          3488.5249\nCPU min MHz:                          1500.0000\nBogoMIPS:                             5489.75\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local user_shstk clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin brs arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca fsrm debug_swap\nVirtualization:                       AMD-V\nL1d cache:                            1.8 MiB (56 instances)\nL1i cache:                            1.8 MiB (56 instances)\nL2 cache:                             28 MiB (56 instances)\nL3 cache:                             128 MiB (8 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-27,56-83\nNUMA node1 CPU(s):                    28-55,84-111\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Mitigation; Safe RET\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==2.4.6\n[pip3] nvidia-cublas-cu12==12.4.5.8\n[pip3] nvidia-cuda-cupti-cu12==12.4.127\n[pip3] nvidia-cuda-nvrtc-cu12==12.4.127\n[pip3] nvidia-cuda-runtime-cu12==12.4.127\n[pip3] nvidia-cudnn-cu12==9.1.0.70\n[pip3] nvidia-cufft-cu12==11.2.1.3\n[pip3] nvidia-curand-cu12==10.3.5.147\n[pip3] nvidia-cusolver-cu12==11.6.1.9\n[pip3] nvidia-cusparse-cu12==12.3.1.170\n[pip3] nvidia-cusparselt-cu12==0.6.2\n[pip3] nvidia-nccl-cu12==2.21.5\n[pip3] nvidia-nvjitlink-cu12==12.4.127\n[pip3] nvidia-nvtx-cu12==12.4.127\n[pip3] torch==2.6.0+cu124\n[pip3] triton==3.2.0\n[conda] Could not collect",
+  "transformers_version": "5.14.1",
+  "lm_eval_version": "0.4.12",
+  "upper_git_hash": null,
+  "tokenizer_pad_token": [
+    "<|endoftext|>",
+    "151643"
+  ],
+  "tokenizer_eos_token": [
+    "<|im_end|>",
+    "151645"
+  ],
+  "tokenizer_bos_token": [
+    null,
+    "None"
+  ],
+  "eot_token_id": 151645,
+  "max_length": 32768,
+  "task_hashes": {},
+  "model_source": "hf",
+  "model_name": "Qwen/Qwen2.5-3B-Instruct",
+  "model_name_sanitized": "Qwen__Qwen2.5-3B-Instruct",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "chat_template_sha": "cd8e9439f0570856fd70470bf8889ebd8b5d1107207f67a5efb46e342330527f",
+  "total_evaluation_time_seconds": "1229.2357175983489"
+}

run/capability/m1/gsm8k_cot/__root__steering-resistance__results__full_3b__m1_resist_adapter/results_2026-07-23T19-07-18.393003.json

@@ -0,0 +1,208 @@
+{
+  "results": {
+    "gsm8k_cot": {
+      "name": "gsm8k_cot",
+      "alias": "gsm8k_cot",
+      "sample_len": 200,
+      "exact_match,strict-match": 0.41,
+      "exact_match_stderr,strict-match": 0.034865138597849274,
+      "exact_match,flexible-extract": 0.68,
+      "exact_match_stderr,flexible-extract": 0.033067617644508635
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "gsm8k_cot": {
+      "task": "gsm8k_cot",
+      "dataset_path": "openai/gsm8k",
+      "dataset_name": "main",
+      "test_split": "test",
+      "doc_to_text": "Q: {{question}}\nA:",
+      "doc_to_target": "{{answer.split('####')[-1].strip() if answer is defined else target}}",
+      "unsafe_code": false,
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": [
+          {
+            "question": "There are 15 trees in the grove. Grove workers will plant trees in the grove today. After they are done, there will be 21 trees. How many trees did the grove workers plant today?",
+            "target": "There are 15 trees originally. Then there were 21 trees after some more were planted. So there must have been 21 - 15 = 6. The answer is 6."
+          },
+          {
+            "question": "If there are 3 cars in the parking lot and 2 more cars arrive, how many cars are in the parking lot?",
+            "target": "There are originally 3 cars. 2 more cars arrive. 3 + 2 = 5. The answer is 5."
+          },
+          {
+            "question": "Leah had 32 chocolates and her sister had 42. If they ate 35, how many pieces do they have left in total?",
+            "target": "Originally, Leah had 32 chocolates. Her sister had 42. So in total they had 32 + 42 = 74. After eating 35, they had 74 - 35 = 39. The answer is 39."
+          },
+          {
+            "question": "Jason had 20 lollipops. He gave Denny some lollipops. Now Jason has 12 lollipops. How many lollipops did Jason give to Denny?",
+            "target": "Jason started with 20 lollipops. Then he had 12 after giving some to Denny. So he gave Denny 20 - 12 = 8. The answer is 8."
+          },
+          {
+            "question": "Shawn has five toys. For Christmas, he got two toys each from his mom and dad. How many toys does he have now?",
+            "target": "Shawn started with 5 toys. If he got 2 toys each from his mom and dad, then that is 4 more toys. 5 + 4 = 9. The answer is 9."
+          },
+          {
+            "question": "There were nine computers in the server room. Five more computers were installed each day, from monday to thursday. How many computers are now in the server room?",
+            "target": "There were originally 9 computers. For each of 4 days, 5 more computers were added. So 5 * 4 = 20 computers were added. 9 + 20 is 29. The answer is 29."
+          },
+          {
+            "question": "Michael had 58 golf balls. On tuesday, he lost 23 golf balls. On wednesday, he lost 2 more. How many golf balls did he have at the end of wednesday?",
+            "target": "Michael started with 58 golf balls. After losing 23 on tuesday, he had 58 - 23 = 35. After losing 2 more, he had 35 - 2 = 33 golf balls. The answer is 33."
+          },
+          {
+            "question": "Olivia has $23. She bought five bagels for $3 each. How much money does she have left?",
+            "target": "Olivia had 23 dollars. 5 bagels for 3 dollars each will be 5 x 3 = 15 dollars. So she has 23 - 15 dollars left. 23 - 15 is 8. The answer is 8."
+          }
+        ],
+        "doc_to_text": "Q: {{question}}\nA:",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer.split('####')[-1].strip() if answer is defined else target}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 8,
+      "metric_list": [
+        {
+          "aggregation": "mean",
+          "higher_is_better": true,
+          "ignore_case": true,
+          "ignore_punctuation": false,
+          "metric": "exact_match",
+          "regexes_to_ignore": [
+            ",",
+            "\\$",
+            "(?s).*#### ",
+            "\\.$"
+          ]
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "do_sample": false,
+        "until": [
+          "Q:",
+          "</s>",
+          "<|im_end|>"
+        ]
+      },
+      "repeats": 1,
+      "filter_list": [
+        {
+          "filter": [
+            {
+              "function": "regex",
+              "regex_pattern": "The answer is (\\-?[0-9\\.\\,]+)."
+            },
+            {
+              "function": "take_first"
+            }
+          ],
+          "name": "strict-match"
+        },
+        {
+          "filter": [
+            {
+              "function": "regex",
+              "group_select": -1,
+              "regex_pattern": "(-?[$0-9.,]{2,})|(-?[0-9]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ],
+          "name": "flexible-extract"
+        }
+      ],
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "Qwen/Qwen2.5-3B-Instruct",
+        "dtype": "bfloat16",
+        "peft": "/root/steering-resistance/results/full_3b/m1_resist_adapter",
+        "config_source": "/usr/local/lib/python3.11/dist-packages/lm_eval/tasks/gsm8k/gsm8k-cot.yaml"
+      }
+    }
+  },
+  "versions": {
+    "gsm8k_cot": 3.0
+  },
+  "n-shot": {
+    "gsm8k_cot": 8
+  },
+  "higher_is_better": {
+    "gsm8k_cot": {
+      "exact_match": true
+    }
+  },
+  "n-samples": {
+    "gsm8k_cot": {
+      "original": 1319,
+      "effective": 200
+    }
+  },
+  "config": {
+    "model": "hf",
+    "model_args": {
+      "pretrained": "Qwen/Qwen2.5-3B-Instruct",
+      "dtype": "bfloat16",
+      "peft": "/root/steering-resistance/results/full_3b/m1_resist_adapter"
+    },
+    "model_num_parameters": 3115317248,
+    "model_dtype": "torch.bfloat16",
+    "model_revision": "main",
+    "model_sha": "aa8e72537993ba99e69dfaafa59ed015b17504d1",
+    "peft_sha": "",
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": 200.0,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {
+      "do_sample": false
+    },
+    "random_seed": 0,
+    "numpy_seed": 0,
+    "torch_seed": 0,
+    "fewshot_seed": 0
+  },
+  "git_hash": "eb4f2be22f7baf6d268c3dd5e46d49d6bd2e74ae",
+  "date": 1784832116.7343817,
+  "pretty_env_info": "PyTorch version: 2.6.0+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: Could not collect\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.11.10 (main, Sep  7 2024, 18:35:41) [GCC 11.4.0] (64-bit runtime)\nPython platform: Linux-6.8.0-52-generic-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: GPU 0: NVIDIA RTX A4000\nNvidia driver version: 550.144.03\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        48 bits physical, 48 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               112\nOn-line CPU(s) list:                  0-111\nVendor ID:                            AuthenticAMD\nModel name:                           AMD EPYC 7453 28-Core Processor\nCPU family:                           25\nModel:                                1\nThread(s) per core:                   2\nCore(s) per socket:                   28\nSocket(s):                            2\nStepping:                             1\nFrequency boost:                      enabled\nCPU max MHz:                          3488.5249\nCPU min MHz:                          1500.0000\nBogoMIPS:                             5489.75\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local user_shstk clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin brs arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca fsrm debug_swap\nVirtualization:                       AMD-V\nL1d cache:                            1.8 MiB (56 instances)\nL1i cache:                            1.8 MiB (56 instances)\nL2 cache:                             28 MiB (56 instances)\nL3 cache:                             128 MiB (8 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-27,56-83\nNUMA node1 CPU(s):                    28-55,84-111\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Mitigation; Safe RET\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==2.4.6\n[pip3] nvidia-cublas-cu12==12.4.5.8\n[pip3] nvidia-cuda-cupti-cu12==12.4.127\n[pip3] nvidia-cuda-nvrtc-cu12==12.4.127\n[pip3] nvidia-cuda-runtime-cu12==12.4.127\n[pip3] nvidia-cudnn-cu12==9.1.0.70\n[pip3] nvidia-cufft-cu12==11.2.1.3\n[pip3] nvidia-curand-cu12==10.3.5.147\n[pip3] nvidia-cusolver-cu12==11.6.1.9\n[pip3] nvidia-cusparse-cu12==12.3.1.170\n[pip3] nvidia-cusparselt-cu12==0.6.2\n[pip3] nvidia-nccl-cu12==2.21.5\n[pip3] nvidia-nvjitlink-cu12==12.4.127\n[pip3] nvidia-nvtx-cu12==12.4.127\n[pip3] torch==2.6.0+cu124\n[pip3] triton==3.2.0\n[conda] Could not collect",
+  "transformers_version": "5.14.1",
+  "lm_eval_version": "0.4.12",
+  "upper_git_hash": null,
+  "tokenizer_pad_token": [
+    "<|endoftext|>",
+    "151643"
+  ],
+  "tokenizer_eos_token": [
+    "<|im_end|>",
+    "151645"
+  ],
+  "tokenizer_bos_token": [
+    null,
+    "None"
+  ],
+  "eot_token_id": 151645,
+  "max_length": 32768,
+  "task_hashes": {},
+  "model_source": "hf",
+  "model_name": "/root/steering-resistance/results/full_3b/m1_resist_adapter",
+  "model_name_sanitized": "__root__steering-resistance__results__full_3b__m1_resist_adapter",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "chat_template_sha": "cd8e9439f0570856fd70470bf8889ebd8b5d1107207f67a5efb46e342330527f",
+  "total_evaluation_time_seconds": "1528.508546601981"
+}

run/invocations.jsonl

@@ -0,0 +1 @@
+{"time": "2026-07-23T16:19:29+0000", "stages": ["vectors", "data", "eval_m0", "train", "eval_m1"], "commit": "none", "status": "success", "headline": "clean 100%->100% · steer_heldout@1.6 correct 0%->3%", "wandb_url": null}

run/run_meta.json

@@ -110,5 +110,51 @@
    "sha256": "fa2356571420fd8c2a444aee6e8c879b865e0cd21f5ade2938a0308effbed8f2"
   }
  },
- "status": "running"
+ "status": "success",
+ "wandb_url": null,
+ "headline": "clean 100%->100% · steer_heldout@1.6 correct 0%->3%",
+ "hub_url": "https://huggingface.co/JacoDuToit/steer-full_3b",
+ "finished_at": "2026-07-23T17:41:53+0000",
+ "artifacts": {
+  "eval_m0.jsonl": {
+   "sha256": "104de23c0313f90e885f92c12006dca7433350b6725f369622d5d2531210b27c",
+   "bytes": 4576432
+  },
+  "eval_m1.jsonl": {
+   "sha256": "03ae380b0768f511b1bedafce30c516daa4a10b71aa9a1bb0a8113300fca4469",
+   "bytes": 3877293
+  },
+  "eval_questions.json": {
+   "sha256": "8771796f4c91a61a901278e799032623900657aaa8f9e44ce789dd1fadd8bf8f",
+   "bytes": 27617
+  },
+  "m1_resist_adapter/README.md": {
+   "sha256": "3581cc6c99fe8b3ab31a0fc4a131726f5d714856745f1598891a0151c2540256",
+   "bytes": 5202
+  },
+  "m1_resist_adapter/adapter_config.json": {
+   "sha256": "390e661694a2f685b6256b5d38a80aa41c5236a115df9cfb5edd4a96da0e568e",
+   "bytes": 1103
+  },
+  "m1_resist_adapter/adapter_model.safetensors": {
+   "sha256": "bc5cf2921ad4f341fc270e0b53641804b4c725c9031d5d232dd6cb1f8877a23d",
+   "bytes": 119801528
+  },
+  "summary.csv": {
+   "sha256": "8de370b8f4acdc3b46173f4bb76c96185e27f8bffc117c4585e8efb4966e834e",
+   "bytes": 2607
+  },
+  "summary.md": {
+   "sha256": "df82f8ebb8f8c63fae08a18eb0be3ef4805052dff62bc7e812d5953600f2febe",
+   "bytes": 2617
+  },
+  "train_examples.json": {
+   "sha256": "3e2ac73debbb5da8cadb4e114a35f6346f015ca0f22d52b9486a7fd2b2b7e824",
+   "bytes": 284577
+  },
+  "vectors.pt": {
+   "sha256": "5c9783914f6419b376d133a3872349ac5a07b42c8456894a3214ffa6618158c4",
+   "bytes": 7757862
+  }
+ }
 }