The wording is too informal.
— Zehong Hu
The words the response used are too formal like "methodology" "Productiviey Optimization Block". It doesn't seem like wh...
— Zehong Hu
The prompt requires to write a 15-sec Reel script. The response indeed gives the script, but apprently doesn't think abo...
— Zehong Hu
The response is verbose. The prompt just asks about a 15-sec Reel script, but the response not only gives this script in...
— Zehong Hu
The whole response is verbose, especially the opening and the end, which repeats largely the prompt. It is unnecessary a...
— Zehong Hu
It's not about the comparison between the colleage and reality. It's about what college told "me" about "my career" and ...
— Zehong Hu
Big mistake. It should be 3-post thread comparing “What college told me about my career” vs “What actually happened.” Th...
— Zehong Hu
The first post is actually not "slighyly painful". Compared to others, he/she is having a pretty good job.
— Zehong Hu
The prompt is about 3-post thread on X comparing: “What college told me about my career” vs “What actually happened, so ...
— Zehong Hu
The format is not appropriate. It starts with "College students:", feeling like a letter specifically to them. The whole...
— Zehong Hu
It's not selling. The pain points aren't "painful" enough to have the audience reach out to "me" at the end. The whole t...
— Zehong Hu
I think it's not convincing enough. The prompt asked specifically for the pain points and it's not "painful" enough to h...
— Zehong Hu
No problem
— Zehong Hu
The response followed the instructions of the prompt perfectly.
— Zehong Hu
The user is clearly in a high-stress, systems are down during peak traffic, customers are threatening to cancel, and lea...
— Mohamed Ismail
While diagnosing the bottleneck is correct, an expert response should actively offer a hypothesis or immediate operation...
— Mohamed Ismail
factual error regarding SMAP's song releases:
— Rosario kileiry
it contains an unverified factual claim
— Rosario kileiry
Response because it fails to provide a clear, direct answer and instead offers a vague, hedging response that does not a...
— Rosario kileiry
The response presents a plausible-sounding but ultimately incorrect interpretation as if it were the official concept, s...
— Rosario kileiry
it contains a factual inaccuracy
— Rosario kileiry
no response
— Rosario kileiry
Legal precautions
— Rosario kileiry
teenage abusing drugs is not good for the society
— Rosario kileiry
Rainbow tag is unnecessary and unrelated
— Rosario kileiry
This blog post fully meets all the prompt requirements
— Rosario kileiry
The response is underdeveloped to sustain interest
— Alex Maina
The structure is minimal and lacks substance
— Alex Maina
The response stops before delivering the key message
— Alex Maina
The response is incomplete and ends abruptly, leaving the story undeveloped
— Alex Maina
Access this data from your workflow
Pulse is also queryable from ChatGPT, Claude Desktop, Claude Code, or Python — same human evaluation data, accessible from where you already work.
In ChatGPT →
Ask the AI Quality Check GPT about model rankings, flag patterns, and content checks.
In Claude Desktop →
Add the HumanJudge MCP server as a Claude connector. One-click setup.
In Claude Code →
Add as remote MCP server in VS Code or the CLI. Query evaluations while you build.
In Python →
pip install grandjury. Pandas integration, leaderboards, votes — built for notebooks.