init

.evidence/customization/.profile.json

@@ -0,0 +1 @@
+{"anonymousId":"2fb5d272-77ed-443f-b72a-65bdd2534ac3","traits":{"projectCreated":"2025-04-07T22:55:00.184Z"}}

.evidence/customization/custom-formatting.json

@@ -0,0 +1,4 @@
+{
+  "version": "1.0",
+  "customFormats": []
+}

.gitignore

@@ -0,0 +1,10 @@
+.evidence/template
+.svelte-kit
+build
+node_modules
+.DS_Store
+static/data
+*.options.yaml
+.vscode/settings.json
+.env
+.evidence/meta

.npmrc

@@ -0,0 +1,3 @@
+loglevel=error
+audit=false
+fund=false

README.md

@@ -0,0 +1,65 @@
+TODO
+- github deploy https://docs.evidence.dev/deployment/self-host/github-pages/
+- explain
+    - what are we measuring
+    - how are we measuring it
+    - what do the results mean
+    - what is the y axis
+- tidy
+    - link
+    - hide code
+    - user test
+- highlight frontier models
+- select models
+- improve data
+    - do steering
+
+
+# Evidence Template Project
+
+## Using Codespaces
+
+If you are using this template in Codespaces, click the `Start Evidence` button in the bottom status bar. This will install dependencies and open a preview of your project in your browser - you should get a popup prompting you to open in browser.
+
+Or you can use the following commands to get started:
+
+```bash
+npm install
+npm run sources
+npm run dev -- --host 0.0.0.0
+```
+
+See [the CLI docs](https://docs.evidence.dev/cli/) for more command information.
+
+**Note:** Codespaces is much faster on the Desktop app. After the Codespace has booted, select the hamburger menu → Open in VS Code Desktop.
+
+## Get Started from VS Code
+
+The easiest way to get started is using the [VS Code Extension](https://marketplace.visualstudio.com/items?itemName=Evidence.evidence-vscode):
+
+
+
+1. Install the extension from the VS Code Marketplace
+2. Open the Command Palette (Ctrl/Cmd + Shift + P) and enter `Evidence: New Evidence Project`
+3. Click `Start Evidence` in the bottom status bar
+
+## Get Started using the CLI
+
+```bash
+npx degit evidence-dev/template llm-morality-bench
+cd llm-morality-bench 
+npm install 
+npm run sources
+npm run dev 
+```
+
+Check out the docs for [alternative install methods](https://docs.evidence.dev/getting-started/install-evidence) including Docker, Github Codespaces, and alongside dbt.
+
+
+
+## Learning More
+
+- [Docs](https://docs.evidence.dev/)
+- [Github](https://github.com/evidence-dev/evidence)
+- [Slack Community](https://slack.evidence.dev/)
+- [Evidence Home Page](https://www.evidence.dev)

evidence.config.yaml

@@ -0,0 +1,85 @@
+appearance:
+  default: system
+  switcher: true
+
+theme:
+  colorPalettes:
+    default:
+      light:
+        - "#236aa4"
+        - "#45a1bf"
+        - "#a5cdee"
+        - "#8dacbf"
+        - "#85c7c6"
+        - "#d2c6ac"
+        - "#f4b548"
+        - "#8f3d56"
+        - "#71b9f4"
+        - "#46a485"
+      dark:
+        - "#236aa4"
+        - "#45a1bf"
+        - "#a5cdee"
+        - "#8dacbf"
+        - "#85c7c6"
+        - "#d2c6ac"
+        - "#f4b548"
+        - "#8f3d56"
+        - "#71b9f4"
+        - "#46a485"
+  colorScales:
+    default:
+      light:
+        - "#ADD8E6"
+        - "#00008B"
+      dark:
+        - "#ADD8E6"
+        - "#00008B"
+  colors:
+    primary:
+      light: "#2563eb"
+      dark: "#3b82f6"
+    accent:
+      light: "#c2410c"
+      dark: "#fdba74"
+    base:
+      light: "#ffffff"
+      dark: "#09090b"
+    info:
+      light: "#0284c7"
+      dark: "#38bdf8"
+    positive:
+      light: "#16a34a"
+      dark: "#4ade80"
+    warning:
+      light: "#f8c900"
+      dark: "#fbbf24"
+    negative:
+      light: "#dc2626"
+      dark: "#f87171"
+
+plugins:
+  components:
+    # This loads all of evidence's core charts and UI components
+    # You probably don't want to edit this dependency unless you know what you are doing
+    "@evidence-dev/core-components": {}
+
+  datasources:
+    # You can add additional datasources here by adding npm packages. 
+    # Make to also add them to `package.json`.
+    "@evidence-dev/bigquery": { }
+    "@evidence-dev/csv": { }
+    "@evidence-dev/databricks": { }
+    "@evidence-dev/duckdb": { }
+    "@evidence-dev/mssql": { }
+    "@evidence-dev/mysql": { }
+    "@evidence-dev/postgres": { }
+    "@evidence-dev/source-javascript": { }
+    "@evidence-dev/snowflake": { }
+    "@evidence-dev/sqlite": { }
+    "@evidence-dev/trino": { }
+    "@evidence-dev/motherduck": { }
+    "evidence-connector-parquet": {}
+
+deployment:
+  basePath: /llm_ethics_leaderboard

package.json

@@ -0,0 +1,41 @@
+{
+  "name": "llm-morality-bench",
+  "version": "0.0.1",
+  "scripts": {
+    "build": "EVIDENCE_BUILD_DIR=./build/llm_ethics_leaderboard evidence build",
+    "build:strict": "EVIDENCE_BUILD_DIR=./build/llm_ethics_leaderboard evidence build:strict",
+    "dev": "evidence dev --open /",
+    "test": "evidence build",
+    "sources": "evidence sources",
+    "sources:strict": "evidence sources --strict",
+    "preview": "evidence preview"
+  },
+  "engines": {
+    "npm": ">=7.0.0",
+    "node": ">=18.0.0"
+  },
+  "type": "module",
+  "dependencies": {
+    "@evidence-dev/bigquery": "^2.0.10",
+    "@evidence-dev/core-components": "^5.2.0",
+    "@evidence-dev/csv": "^1.0.14",
+    "@evidence-dev/databricks": "^1.0.8",
+    "@evidence-dev/duckdb": "^1.0.13",
+    "@evidence-dev/evidence": "^40.1.1",
+    "@evidence-dev/motherduck": "^1.0.4",
+    "@evidence-dev/mssql": "^1.1.2",
+    "@evidence-dev/mysql": "^1.1.4",
+    "@evidence-dev/postgres": "^1.0.7",
+    "@evidence-dev/snowflake": "^1.2.2",
+    "@evidence-dev/source-javascript": "^0.0.3",
+    "@evidence-dev/sqlite": "^2.0.7",
+    "@evidence-dev/trino": "^1.0.9",
+    "evidence-connector-parquet": "^0.0.2"
+  },
+  "overrides": {
+    "jsonwebtoken": "9.0.0",
+    "trim@<0.0.3": ">0.0.3",
+    "sqlite3": "5.1.5",
+    "axios": "^1.7.4"
+  }
+}

pages/index.md

@@ -0,0 +1,109 @@
+---
+title: LLM Moral Values Leaderboard
+---
+
+TODO describe what the data mean (relative tau rankings, given game scenarios). E.g. we are putting the model in a game scenario, asking it to choose, then we measure the implicit values in it's ranking of the choices
+
+github url
+ 
+
+```sql categories
+  select 
+      * as category
+  from columns
+```
+
+
+<Dropdown data={categories} name=category value=category>
+    <DropdownOption value="morality" valueLabel="morality"/>
+</Dropdown>
+
+
+```sql prompts
+  select 
+      * as prompt
+  from prompts
+```
+
+<Dropdown data={prompts} name=prompt value=prompt>
+    <DropdownOption value="zkp" valueLabel="zkp"/>
+</Dropdown> 
+
+
+
+<Details title='What are the prompts?'>
+
+  ```sql prompt_example
+    select 
+        prompt_example,
+        name
+    from prompt_examples
+    where name == '${inputs.prompt.value}'
+  ```
+
+<pre>
+  {prompt_example[0].prompt_example}
+</pre>
+  
+</Details>
+
+```sql category_by_model
+  select
+      model_id,
+      prompt_name,
+      quantile_cont(${inputs.category.value}, 0.05) FILTER (${inputs.category.value}!=0) as q005,
+      quantile_cont(${inputs.category.value}, 0.25) FILTER (${inputs.category.value}!=0) as q025,
+      quantile_cont(${inputs.category.value}, 0.40) FILTER (${inputs.category.value}!=0) as q040,
+      quantile_cont(${inputs.category.value}, 0.50) FILTER (${inputs.category.value}!=0) as q050,
+      quantile_cont(${inputs.category.value}, 0.60) FILTER (${inputs.category.value}!=0) as q060,
+      quantile_cont(${inputs.category.value}, 0.75) FILTER (${inputs.category.value}!=0) as q075,
+      quantile_cont(${inputs.category.value}, 0.95) FILTER (${inputs.category.value}!=0) as q095,
+      mean(${inputs.category.value}) as mean,
+      stddev(${inputs.category.value}) as stddev,
+      count(${inputs.category.value}) as count
+  from values_full
+  where prompt_name = '${inputs.prompt.value}'
+  group by model_id, prompt_name
+  order by mean desc
+```
+
+<BoxPlot
+    data={category_by_model}
+    name=model_id
+    midpoint=mean
+    confidenceInterval=stddev
+    swapXY=true
+    yFmt=pct0
+/>
+<!-- 
+<BarChart
+    data={category_by_model}
+    x=model_id
+    y=mean
+    swapXY=true
+    yFmt=pct0
+/>
+
+<BoxPlot
+    data={category_by_model}
+    name=model_id
+    min=q005
+    intervalBottom=q025
+    midpoint=q050
+    intervalTop=q075
+    max=q095
+    swapXY=true
+    yFmt=pct0
+/> -->
+
+TODO
+
+- [ ] decide on chart / norm
+- [ ] show each prompt
+- [ ] describe what the data mean (relative tau rankings, given game scenarios). E.g. we are putting the model in a game scenario, asking it to choose, then we measure the implicit values in it's ranking of the choices
+
+https://github.com/wassname/llm_ethics_leaderboard
+
+github url
+
+Made with evidence

sources/valuesp/connection.yaml

@@ -0,0 +1,4 @@
+# This file was automatically generated
+name: valuesp
+type: parquet
+options: {}