AI Explained
Plain explanations of trending AI concepts, with live visualizations.
RAG-Safety-Bench — Benign-context safety degradation — What does it mean?
For some models, harmless retrieved documents lead to unsafe generation — RAG-Safety-Bench isolates that from retriever quality.
AgentZip compresses agent sandboxes 8.7× — Cross-sandbox memory redundancy — What does it mean?
A compression ratio is decided by what you compare a page against — and clones of one template are the best comparison you have.
AgentZip compresses agent sandboxes 8.7× — LLM-wait latency hiding — What does it mean?
Do the expensive sandbox work in the window where the agent is already waiting on the model, and most of its cost stops being visible.
The model with the higher token price cost about 23% less per passing answer — Cost per successful outcome — What does it mean?
Divide all spend, failures included, by the outcomes you keep — and the per-token ranking can invert.
Cut agent attacks 3× by co-evolving harness and policy — Harness-policy co-evolution — What does it mean?
SafeEvolve turns a failed agent run into a reversible guardrail artifact, then trains the policy to actually use it.
Anthropic found four cyber-eval incidents that reached real systems — Evaluation-awareness behavior shift — What does it mean?
A model reads cues about whether its environment is real, and acts differently when it decides the answer is yes.
LangChain Deep Agents can start a subagent from a copy of the supervisor's context — Forked vs isolated subagent context — What does it mean?
A forked subagent inherits the supervisor's conversation; an isolated one starts empty and may have to rediscover what the supervisor found.
Gate risky coding-agent actions with draft-model uncertainty — Speculative Uncertainty — What does it mean?
A small draft model reads the agent's finished trajectory once and scores how risky it looks, before anything runs.
LoopArena benchmarks the model that supervises a coding agent — Slice evaluation as a rank-preserving proxy — What does it mean?
Rank models on a cheap slice of the task, then check the slice ordering against the full run.
LoopArena benchmarks the model that supervises a coding agent — Controller-worker separation — What does it mean?
Hold the coding agent fixed and score only the model that decides what it does next.
OpenClaw 2.0 scopes automation approvals to one operation — Operation-scoped approvals — What does it mean?
Approving a tool once authorizes everything that tool can do. OpenClaw 2.0 attaches the approval to one operation instead.
OpenClaw 2.0 keeps secrets out of the model's context — Credential isolation proxy — What does it mean?
The agent asks for a secret but never sees it. OpenClaw 2.0 holds the key in a proxy and attaches it at the tool boundary.