lexi

[Miami Marlins @Marlins]

GRIFFIN CONINE. BALLGAME.

[Elon Musk @elonmusk]

Grok 4.5 true usefulness is excellent

[Snorkel AI @SnorkelAI]

We’re launching Senior SWE-Bench today: a new open benchmark that evaluates AI coding agents as senior engineers. Realistic instructions, long horizons, and code quality assessments. Built at Snorkel AI in collab with @Princeton and @UWMadison.

Claude Opus 4.8 is the leader with 24% high quality solves, but it needed 117K tokens on average to get there. Claude Sonnet 5 almost swooped in for the win, but we found it cheated on 26% of trials (analysis on the site). GPT-5.5 is the most efficient high performer, using only 36K tokens at its highest effort level to land at #2.

Led by @henryehrenberg with @vincentsunnchen (Snorkel); @karthik_r_n and Austin W. Hanjie (@Princeton); @gorlanski and @fredsala (@UWMadison).

[Henry Kiss Ehrenberg @henryehrenberg]

We expect agents to act like senior engineers, but most benchmarks still evaluate them like interns.

Excited to introduce Senior SWE-Bench, an open-source and @harborframework-native benchmark that assesses agents as senior engineers on long-horizon tasks with realistically under-specified instructions.

We expect agents to build real features going on just a quick Slack message, nothing like the super technical instructions most benchmarks provide. Senior SWE-Bench fixes that.

Claude Opus 4.8 is the current leader at 24% high quality solves, but it took 117K tokens on average to get there. Claude Sonnet 5 looked like it was going to swoop in for the top spot, but we found it cheated on 26% of trials.

[vincent sunn chen @vincentsunnchen]

https://t.co/iiCZcXT9Dg

[Snorkel AI @SnorkelAI]

Only 2 days until we’re in London for @aiDotEngineer 🇬🇧

Come connect with the Snorkel AI team and fellow leaders shaping the future of AI agents – over drinks and conversation 🍻

Last chance to register for Tuesday’s Happy Hour at Bantof: https://t.co/aZmb6vWPGW

[Snorkel AI @SnorkelAI]

London 🇬🇧

We’re heading to @aiDotEngineer Europe (April 8–10). @vincentsunnchen will be presenting on the art and science of benchmarking agents, including learnings from Open Benchmark Grants.

Find us at Booth G9:
https://t.co/VouHA6Ep8h

[vincent sunn chen @vincentsunnchen]

https://t.co/aNuEf6Yu9j

[Miami Marlins @Marlins]

Conine 🤝⚾️ Conine! It’s a Miami Marlins family affair with tonight’s first pitch.
#homeofbeisbol

[Barstool Florida @UFBarstool]

STRAIGHT INTO MY VEINS @GatorsFB

[Miami HEAT @MiamiHEAT]

Nobody does Media Day like Jimmy Butler does Media Day

[𝙃𝙀𝘼𝙏 𝙉𝘼𝙏𝙄𝙊𝙉 @HeatvsHaters]

10 YEARS AGO TODAY

Rebound Bosh.
Back out to Allen.
His three-pointer…..

[Miami HEAT @MiamiHEAT]

We made history, don’t let them tell you differently. Thank you #HEATNation for always riding with us and FOR us!

[lexi @alexissobel]

Miss you already @ThisIsUD ❤️

[lexi @alexissobel]

Don’t let us get one, @nuggets 😉

[Bleacher Report @BleacherReport]

Miami's UNREAL playoff run continues

-Lost first play-in game to Hawks
-Beat Bulls in play-in finale for No. 8 seed
-Knocked out No. 1 seed Bucks in 5
-Beat No. 5 seed Knicks in 6
-Beat No. 2 seed Celtics in ECF
-First play-in team to make the Finals
-Second No. 8 seed to ever make the Finals

[Brady Hawk @BradyHawk305]

History was indeed made

The 8 seeded Miami Heat are heading to the NBA Finals

[𝙃𝙀𝘼𝙏 𝙉𝘼𝙏𝙄𝙊𝙉 @HeatvsHaters]

LEAD US, Highsmith

[𝙃𝙀𝘼𝙏 𝙉𝘼𝙏𝙄𝙊𝙉 @HeatvsHaters]

Butler’s timeout signal to Horford.

LOVE THAT MAN 😭😭

[lexi @alexissobel]

@MiamiHEAT LEAD US CALEB!!!!!!!!!