Results for “heckler-handling”

Runs directional ablation and refusal-direction analysis for open-weight models the user may modify; use to reduce benign over-refusal or measure refusal/KL trade-offs, not for training.

Manage Hugging Face Hub resources: download/upload models, datasets, spaces; manage repos, buckets, collections, discussions, and cache; run SQL queries on datasets; authenticate and manage tokens.

Train or fine-tune language and vision models using TRL or Unsloth on Hugging Face Jobs cloud infrastructure, with support for SFT, DPO, GRPO, and reward modeling, plus GGUF conversion for local deployment.

Persistent task ledger for agent coordination. Plan multi-step work, checkpoint progress across session boundaries, and coordinate across multiple agents with project pool routing.

Package an in-progress task on this cluster so a Claude agent on another cluster (MMM Young, CX3, HX1, Isambard) can pick it up, or resume a task handed off to this cluster. Use when the user says hand off / migrate / continue this on <cluster>, or asks you to pick up an existing handoff.

Use to analyse average handle time, resolution time or time-in-queue without being misled by the skew, and to find where time actually goes. Trigger for "why is our AHT increasing", "our call handling times are up", "which contact reasons take longest", "how long do tickets take", handle time by agent or team, or an AHT target being set.

Design and verify auditable human oversight, approval gates, escalation paths, and safe state transitions for AI agent workflows. Use when deciding which agent actions require review, adding approve/reject or dual-control flows, preventing unauthorized autonomous effects, creating decision records, reducing rubber-stamping, or recovering safely from rejected, expired, or failed actions.

Offload sub-tasks to other agents, specialist models, or background jobs, with guidance on choosing the right method and verifying results.

Hands the current conversation to a fresh background agent that picks up the work immediately. Use when the user wants the work continued by a background agent.

Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware, with backend selection between vLLM, Transformers, and accelerate.

Compact the current conversation into a handoff document for another agent to pick up, including suggested skills and references to existing artifacts.

Hand the current conversation off to a fresh background agent that picks up the work immediately.

Manage Hugging Face Hub resources via the `hf` CLI: download and upload models, datasets, and spaces; manage buckets, cache, collections, discussions, and inference endpoints; run SQL queries on datasets.

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

Compacts the current conversation into a handoff document for another agent to pick up, referencing existing artifacts by path or URL instead of duplicating them.

Extend and upgrade Hermes Agent's context compression system — StagedArchiver, knowledge fingerprinting, /uncompress command, look-ahead triggers, and schema migration patterns.

End-of-session context handoff. Saves session state so the next session or a parallel agent can pick up exactly where the left off. Run this when context is getting full (~80%) or before stopping work.

Compact the current conversation into a handoff document for another agent to pick up.