AI Paper Atlas

Research map / Adversarial robustness and security

Backdoor and poisoning: research map

385 papers on Backdoor and poisoning in Adversarial robustness and security, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 5 approaches. The busiest year so far is 2026.

What it is

An attacker who can tamper with training data can plant a backdoor: the model behaves normally except when a secret trigger appears. This topic studies such attacks, including on large and federated models, and defences that detect poisoned data or remove backdoors.

Terms you will see on this page:

Backdoor
hidden behaviour activated by a trigger pattern.
Data poisoning
corrupting training data to change what a model learns.
Trigger
the pattern, such as a small patch or phrase, that activates the backdoor.

Start reading with the most cited papers here:

Within Adversarial robustness and security, its share grew from 13.8% in 2023–24 to 15.2% in 2025–26 (107 → 149 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).

2016: 2162017: 1172018: 2182019: 1192020: 12202021: 23212022: 42222023: 53232024: 74242025: 80252026: 9526

Explore Backdoor and poisoning in the interactive map

Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.

What changed recently

Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):

Approaches and key papers

Backdoor attacks · 275 papers
Keywords: backdoor attack, poisoning, triggers

Approaches in this cluster:

Most cited and most cited since 2024:

Backdoor defenses · 110 papers
Keywords: backdoor defense, backdoored, clean

Approaches in this cluster:

Most cited and most cited since 2024:

Related topics in Adversarial robustness and security

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy