Google 🇺🇸 · Gemma 3

Gemma 3 4B IT

An instruction-tuned 4B multimodal Gemma 3 checkpoint with function calling and long context.

DOOM SCORE51.7out of 100model risk profile, not the overall index
0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 3

Why this model scores 51.7

Instruction following, vision and function calling raise capability and autonomy. Small open weights maximize diffusion and make downstream controls hard to enforce.

Capability49
Autonomy14
Deployment100
Misuse potential45
Control difficulty62
MODEL-ATTRIBUTED EVIDENCE

News tied to Gemma 3 4B IT

The model score of 51.7 rates this model's risk profile. The overall Doom Index of 67.9 measures the complete temporally weighted evidence record. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.02

Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.

AUDIT TRAIL

Model score history

  1. R3
    Doom Score 51.7

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-142416 under temporal-monthly-pressure-v4.

    12 Aug 2026
  2. R2
    Doom Score 51.9

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-132112 under fixed-sensitivity-v3.

    12 Aug 2026
  3. R1
    Doom Score 52.6

    Initial exact profile for the separately released 4B instruction checkpoint.

    12 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for Gemma 3 4B IT.
  1. Gemma 3 4B IT by Google has a DoomBench model risk score of 51.7 out of 100, based on five transparent version-specific dimensions rather than the overall index.

  2. Gemma 3 4B IT's highest current DoomBench dimension is deployment at 100.0 out of 100; the profile publishes every component score and its editorial rationale.

  3. DoomBench links 1 source-backed evidence item to Gemma 3 4B IT, while keeping the model's risk profile separate from each item's contribution to the live Doom Index.

    https://www.doombench.com/models/google-deepmind-gemma-3-4b-it