Skip to content
Longterm Wiki
Index
Grant·ym8kEipMJs·Record·Profile

Grant: Meridian — Avoiding Encoded Reasoning in LLMs (Coefficient Giving → Meridian)

Verdictconfirmed95%
1 check · 4/9/2026

Deterministic match: grantee, amount, date matched in source snapshot (2714 rows)

Our claim

entire record
Name
Meridian — Avoiding Encoded Reasoning in LLMs
Amount
$244,614
Currency
USD
Date
June 2025
Notes
[Navigating Transformative AI] Open Philanthropy recommended a grant of $244,614 to Meridian to support research on maintaining faithfulness in the chain of thought of LLMs. Many researchers are concerned about scenarios where LLM chains of thought might start to hide intermediaexpand[Navigating Transformative AI] Open Philanthropy recommended a grant of $244,614 to Meridian to support research on maintaining faithfulness in the chain of thought of LLMs. Many researchers are concerned about scenarios where LLM chains of thought might start to hide intermediate reasoning steps, encoding these thoughts in ways that human readers won’t detect. This study will evaluate several methods for avoiding encoded reasoning. This grant was funded via a request for proposals for projects related to technical AI safety research. This falls within our focus area of potential risks from advanced artificial intelligence.

Source evidence

1 src · 1 check
confirmed95%deterministic-row-match · 4/9/2026
Name
Meridian — Avoiding Encoded Reasoning in LLMs
Grantee
Meridian
Focus Area
Navigating Transformative AI
Amount
$244,614.00
Date
June 2025
Description
Open Philanthropy recom

NoteDeterministic match: grantee, amount, date matched in source snapshot (2714 rows)

Case № ym8kEipMJsFiled 4/9/2026Confidence 95%
Source Check: Grant: Meridian — Avoiding Encoded Reasoning in LLMs (Coefficient Giving -> Meridian) | Longterm Wiki