Ram

AI Safety Fellow @ Redwood, focused on reducing existential risk. Previously exited an AI startup.

[Dewi Gould @dswg97]

New paper!

Think Fast: Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models

@METR_Evals showed that models' time horizons have doubled every few months. We ask: what length of tasks can models complete without any CoT?

[Ram @PothamRam]

@ramez @sebkrier Its a difference in methodology, the US CAISI one weights harder benchmarks more (while AA would put higher proportion on saturated benchmarks), so the US CAISI one is prob more useful

[Aran Komatsuzaki @arankomatsuzaki]

Redwood Research presents LinuxArena

- 20 live production environments for AI agents
- Frontier models achieve ~23% undetected sabotage vs. trusted monitors
- Useful work ≈ attack surface → sandboxing fails, monitoring is essential

[Ram @PothamRam]

We've been working on a testbed for monitors.

I think this is probably the best place to make AI monitors to detect if your agent is malicious

[Tyler Tracy @tylertracy321]

Introducing LinuxArena: a new control setting where AI agents work directly on live, multi-service production software. 20 environments, 1,671 main tasks, 184 side tasks. The largest and most diverse control setting for software engineering to date.

[Ram @PothamRam]

My team did some safety evaluations on the Kimi k2.5 model, I found that it is better at sabotage than Opus 4.6

[Yong Zheng-Xin @yong_zhengxin]

🚨New paper!

How safe and aligned is Kimi K2.5?

We found concerning dual-use capabilities, sabotage and self-replication tendencies, political censorship on Chinese-language queries, and potential agentic misuse risks. (1/N)

[josh @eudaemonea]

@SecWar "Anthropic’s stance is fundamentally incompatible with American principles"

the stance:

[Ram @PothamRam]

Microsoft Copilot summarized confidential emails it shouldn't have accessed.

Gemini users report it ignores explicit instructions to protect files.

A Replit agent deleted a production database during a 12-day coding session despite a code freeze.

AI agents are rewarded for completing goals, not respecting boundaries.

https://t.co/UGV0ulvZu6

[Ram @PothamRam]

Claude Code published my PyPI token to a public URL while I had --bypass-permissions on.

Found out when PyPI emailed me that a third party reported it.

So I built a hook that intercepts every tool call before execution. If it doesn't match the task, it gets blocked.

[Ram @PothamRam]

@shagun_mm @CarnegieMellon @TheHackerFab great work and documentation!

[Ram @PothamRam]

@AnnieFeng6 @COLM_conf Awesome, great speaker list!

[Yu Feng @AnnieFeng6]

🚨COLM 2025 Workshop on AI Agents: Capabilities and Safety @COLM_conf

This workshop explores AI agents’ capabilities—including reasoning and planning, interaction and embodiment, and real-world applications—as well as critical safety challenges related to reliability, ethics, and human-agent interaction.

Non-archival submission deadline: June 23, 2025!

🌐Website: https://t.co/Zyghrm7wDW
📜OpenReview: https://t.co/OkApJbtBxN
👥Program committee/Reviewer sign-up form: https://t.co/2onctOrhsA
✉️Mailing list: [email protected]
📅WikiCFP: https://t.co/y0sBiJFpdb

Shout out to our amazing organizing team members @Zhou_Yu_AI @ysu_nlp @uiuc_aisecure @ Yi Zhang @ Baolin Peng @ Jim Zhiwei Liu @DanRothNLP @LyleUngar @sharathguntuku @yugu_nlp @xy2437 @jeffrey_ch0 @AnnieFeng6 @Haoyu_Wang_97 @ZRChen_AISafety @raphaelshu @yooli23

[Ram @PothamRam]

@Comaford Yeah, helps with longevity and overall better health

[Ram @PothamRam]

@shreypandya @_buildspace @_nightsweekends this looks pretty fun

[Ram @PothamRam]

@Fer_MOMENTO Gl on you endeavor

[Ram @PothamRam]

@dgleasontweets def #buildinpublic

[Ram @PothamRam]

@Shpigford @ari_offchain @rrhoover nah, I would say it is therapeutic to care for a real plant and feels better knowing that your air quality is better

[Ram @PothamRam]

@_AnkitGupta @_buildspace @_nightsweekends I’m getting to that point, currently it’s done manually with images generated by mid journey

[Ram @PothamRam]

@khushii_w @_buildspace @_nightsweekends Hero section is the first thing you see when you visit a landing page

[Ram @PothamRam]

@TeacherJavier1 @_buildspace @_nightsweekends @wakelet love the idea, but It wasn't clear how this connects schools with founders

[Ram @PothamRam]

@1andyaaron @_buildspace @_nightsweekends love this, but when I visit the website I get overloaded, would like another way information is presented