Skip to content
Navigation menu
Search
Powered by Algolia
Search
Log in
Create account
DEV Community
Close
Receipts, Not Scores Series' Articles
Back to Reid Marlow's Series
Your RAG Eval Is Checking the Receipt, Not the Patient
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Jul 13
Your RAG Eval Is Checking the Receipt, Not the Patient
#
ai
#
rag
#
llm
#
programming
1
reaction
Comments
2
comments
3 min read
Moderation
Agent Debugging Needs More Than Traces
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Jul 23
Agent Debugging Needs More Than Traces
#
ai
#
programming
#
agents
#
debugging
2
reactions
Comments
5
comments
5 min read
Moderation
LLM-as-a-Judge Is Too Expensive to Be the Default
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Jul 28
LLM-as-a-Judge Is Too Expensive to Be the Default
#
ai
#
machinelearning
#
llm
#
agents
4
reactions
Comments
6
comments
4 min read
Moderation
Agent-Built Software Still Needs a Human-Shaped Test
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Jul 31
Agent-Built Software Still Needs a Human-Shaped Test
#
ai
#
programming
#
opensource
#
webdev
2
reactions
Comments
Add Comment
3 min read
Moderation
OpenAI's Math Post Is Really About Audit Trails
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Aug 1
OpenAI's Math Post Is Really About Audit Trails
#
ai
#
agents
#
devtools
#
math
2
reactions
Comments
6
comments
4 min read
Moderation
The coding agent harness paper finally ran component ablations
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Sep 18
The coding agent harness paper finally ran component ablations
#
ai
#
agents
#
coding
#
softwareengineering
5
reactions
Comments
2
comments
3 min read
Moderation
Multi-Agent Debate Sharpens the Explanation, Not the Decision
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Sep 25
Multi-Agent Debate Sharpens the Explanation, Not the Decision
#
ai
#
agents
#
devtools
#
programming
7
reactions
Comments
9
comments
3 min read
Moderation
When Agent Evals Score Cash Balance, the Model Invents Refund Fraud
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Oct 4
When Agent Evals Score Cash Balance, the Model Invents Refund Fraud
#
discuss
#
ai
#
testing
#
programming
9
reactions
Comments
13
comments
4 min read
Moderation
Agents Score 97% on Static Tool Judgments and Still Break Interactive Workflows
Reid Marlow
Reid Marlow
Reid Marlow
Follow
Oct 7
Agents Score 97% on Static Tool Judgments and Still Break Interactive Workflows
#
ai
#
agents
#
automation
#
programming
7
reactions
Comments
13
comments
7 min read
Moderation
We're a place where coders share, stay up-to-date and grow their careers.
Log in
Create account