teleo/teleo-codex
8
0
Fork 0
Code Issues 22 Pull requests 85 Projects Releases Packages Wiki Activity Actions 74

theseus: extract from 2025-11-00-operationalizing-pluralistic-values-llm-alignment.md

Browse source 
- Source: inbox/archive/2025-11-00-operationalizing-pluralistic-values-llm-alignment.md
- Domain: ai-alignment
- Extracted by: headless extraction cron (worker 5)

Pentagon-Agent: Theseus <HEADLESS>
This commit is contained in:
Teleo Agents 2026-03-12 07:51:48 +00:00
parent ba4ac4a73e
commit 14eb6e701e
1 changed files with 14 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
inbox/archive/2025-11-00-operationalizing-pluralistic-values-llm-alignment.md
View file

@ -7,9 +7,14 @@ date: 2025-11-01
domain: ai-alignment domain: ai-alignment
secondary_domains: [] secondary_domains: []
format: paper format: paper
status: unprocessed status: null-result
priority: high priority: high
tags: [pluralistic-alignment, demographic-composition, empirical, safety-inclusivity, real-human-feedback] tags: [pluralistic-alignment, demographic-composition, empirical, safety-inclusivity, real-human-feedback]
processed_by: theseus
processed_date: 2026-03-11
enrichments_applied: ["community-centred norm elicitation surfaces alignment targets materially different from developer-specified rules.md", "pluralistic alignment must accommodate irreducibly diverse values simultaneously rather than converging on a single aligned state.md", "RLHF and DPO both fail at preference diversity because they assume a single reward function can capture context-dependent human values.md"]
extraction_model: "anthropic/claude-sonnet-4.5"
extraction_notes: "First large-scale empirical study quantifying demographic composition effects on alignment outcomes. Single new claim extracted with high confidence (likely) due to large N and systematic design. Three enrichments to existing pluralistic alignment claims. Full paper access would enable extraction of interaction effects and comparison with PAL/MixDPO approaches."
--- ---
## Content ## Content
@ -37,3 +42,11 @@ Demonstrates that "whose feedback" matters as much as "how much feedback" for al
PRIMARY CONNECTION: community-centred norm elicitation surfaces alignment targets materially different from developer-specified rules PRIMARY CONNECTION: community-centred norm elicitation surfaces alignment targets materially different from developer-specified rules
WHY ARCHIVED: Empirical evidence that "whose preferences" is a quantitatively important question, not just a fairness concern WHY ARCHIVED: Empirical evidence that "whose preferences" is a quantitatively important question, not just a fairness concern
EXTRACTION HINT: Focus on the magnitude of demographic composition effects and what this means for single-population alignment training EXTRACTION HINT: Focus on the magnitude of demographic composition effects and what this means for single-population alignment training
## Key Facts
- Study included 1,095 participants providing 27,375 ratings
- Liberal feedback models: +5.0 pp vs Conservative baseline
- White feedback models: +4.7 pp vs Black baseline
- Female feedback models: +3.4 pp vs Male baseline
- Effects measured on emotional awareness and toxicity dimensions