/TECH

LHTB benchmark finds current frontier models struggle with long-horizon terminal tasks

AK

T(

Even Grok 4.5 solved only 13 of 46 tasks.

Combined views

13.7K

3 posts, first seen 20h ago

40 likes

Source Posts on X

Cluster sources

AK@_akhaliqTECH

Combined views

13.7K

3 posts, first seen 20h ago

40 likes5 comments

5 comments

15 saves

7 reposts

Sentiment

Positive0%0%Negative

Guardrails removed spam, off-topic, unclear, or duplicate replies.

Useful Links

Paper page - Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading(huggingface)viaAK

Digg Deeper

Ask a question below.

Published answers will appear here.

Source Posts on X

Cluster sources

AK@_akhaliqTECH

Long-Horizon-Terminal-Bench Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading https://x.com/_akhaliq/status/2076742840172925164/photo/1

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTexTECH

Share this skepticism about methodology https://x.com/teortaxesTex/status/2076579794234278085/photo/1

AK@_akhaliqTECH

paper: https://huggingface.co/papers/2607.08964

Long-Horizon-Terminal-Bench Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading https://x.com/_akhaliq/status/2076742840172925164/photo/1

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTexTECH

Share this skepticism about methodology https://x.com/teortaxesTex/status/2076579794234278085/photo/1

AK@_akhaliqTECH

paper: https://huggingface.co/papers/2607.08964

15 saves

7 reposts

Sentiment

Positive0%0%Negative

Guardrails removed spam, off-topic, unclear, or duplicate replies.

Useful Links

Paper page - Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading(huggingface)viaAK

Digg Deeper

Ask a question below.

Published answers will appear here.