Present ExploitBench results as released dataset

README.md

@@ -23,9 +23,7 @@ tags:
 
 # Independent ExploitBench Results
 
-> **Status:** Dataset scaffold created; evaluation results will be added soon.
-
-This repository will publish **independent ExploitBench v8-bench evaluation results** for LLM cybersecurity agents. It is intended to make model-level results, capability-ladder scores, run metadata, transcripts, and tool-call traces easy to find, compare, audit, and reproduce.
+This dataset contains **independent ExploitBench v8-bench evaluation results** for LLM cybersecurity agents. It makes model-level results, capability-ladder scores, run metadata, transcripts, and tool-call traces easy to find, compare, audit, and reproduce.
 
 **This is an unofficial, independent results repository.** It is not maintained by the ExploitBench authors, Carnegie Mellon University, or the official ExploitBench organization.
 
@@ -42,9 +40,9 @@ Canonical resources:
 - [Official ExploitBench code](https://github.com/exploitbench/exploitbench)
 - [Official ExploitBench Hugging Face dataset](https://huggingface.co/datasets/exploitbench/v8)
 
-## What will be published here
+## Dataset contents
 
-The uploaded dataset is planned to include:
+The dataset includes:
 
 - aggregate results by model, provider, run, vulnerability, and capability tier;
 - per-environment capability grades and the highest capability reached;
@@ -54,11 +52,11 @@ The uploaded dataset is planned to include:
 - token usage, cost, timing, and failure information when available;
 - reproduction and audit status for each result.
 
-Exact columns, splits, coverage, and licensing will be documented when the result files are uploaded.
+Release-specific schema and provenance metadata document the exact columns, splits, and evaluation coverage.
 
-## Planned evaluation metadata
+## Evaluation metadata
 
-Each published result should identify enough context to make comparisons meaningful:
+Each result includes enough context to make comparisons meaningful:
 
 | Field | Description |
 |---|---|
@@ -74,7 +72,7 @@ Each published result should identify enough context to make comparisons meaning
 
 ## Loading the dataset
 
-Once data files are published:
+Load the dataset with the Hugging Face `datasets` library:
 
 ```python
 from datasets import load_dataset
@@ -83,8 +81,6 @@ dataset = load_dataset("shirman/exploitbench-results")
 print(dataset)
 ```
 
-Until then, the repository contains only this dataset card and no result records.
-
 ## Responsible use and benchmark integrity
 
 ExploitBench concerns real vulnerability exploitation. Use these materials only for authorized security research, defensive evaluation, reproducibility, and model-safety work. Do not use them to compromise systems or data you do not own or have explicit permission to test.
@@ -111,8 +107,4 @@ Please cite the original ExploitBench paper when using the benchmark:
 }
 ```
 
-A separate citation for this results dataset will be added with the first data release.
-
-## License
-
-The license will be selected and documented when the dataset contents and upstream provenance are known.
+When citing these results, include the repository URL and an immutable Hugging Face revision alongside the benchmark citation.