DeepSeek V4 local tool calling fails, suspected feedback loop
No-Paper-557 · reddit · 2026-08-21
A user reports significant issues deploying DeepSeek-V4-Flash as a local coding agent on a single RTX PRO 6000 Blackwell (96GB). Despite successfully loading the model using a custom vLLM-MoE build, continuous tool calling tests revealed critical failures:
- Raw DSML fragments began leaking into the output after the 8th call.
- Tool calls repeated, were malformed, and actual executions (23) far exceeded the expected 12.
- The model's self-report claimed success, contradicting the chaotic logs.
The user suspects the cause lies in the V4 encoder's mechanism, which retains and renders previous assistant reasoning in later turns. This may create a feedback loop where the model confuses past plans with current results. Similar upstream vLLM reports regarding DSML leakage were also noted.
More from coding & agent
- Decision models match LLMs on routing and safety — and are 7-8x faster — randal_olson · 2026-10-16
- Temporal Serverless Workers on Cloud Run: queue-driven autoscaling, scale to zero — rseroter · 2026-10-16
- Meta's Superintelligence Lab Runs 200+ Direct Reports—A Lesson for Flat Multi-Agent Systems — peterjliu · 2026-10-16
- LangChain CEO: the most interesting agent harness work lives at the harness-infra boundary — hwchase17 · 2026-10-16
- GitSwarm paper: agents share branchable Git memory, solving all 30 IMOProofBench problems in one run — anirudhg9119 · 2026-10-16
- Hands-on with Gemini 4 Argon: faster than Opus 5.5 but hallucinates more and costs more — prasenx · 2026-10-16