---
title: 'TESSERA v2-1B-M: Distilled EO Foundation Model'
url: https://www.emergentmind.com/topics/tessera-v2-1b-m
type: topic
---

# TESSERA v2-1B-M: Distilled EO Foundation Model

TESSERA v2-1B-M is the medium-capacity distilled deployment model in the TESSERA v2 family of pixel-wise Earth-observation foundation models. It is distilled from a frozen 1B teacher trained on annual Sentinel-1 and Sentinel-2 time series, emits a 128-dimensional embedding with Matryoshka-compatible prefixes at $d\in\{16,32,64,128\}$, and is positioned as “a balanced default” for embeddings-as-data deployment. On the reported 29-task benchmark suite, it achieves the highest composite score among the systems compared, at $0.611$, exceeding AlphaEarth, TESSERA v1, and OlmoEarth-L [2607.03949].

## 1. Family position and nomenclature

The exact identifier **TESSERA v2-1B-M** is defined by the TESSERA v2 scaling paper. In that paper, the model is described as the **21M-parameter student** distilled from a **1B-parameter pixel-wise teacher**, with the suffix **“M”** denoting the **medium-capacity student** in the deployment ladder [2607.03949]. By contrast, the original TESSERA paper established the broader TESSERA system as a dual-encoder remote sensing foundation model producing annual 10 m, 128-dimensional pixel embeddings, but it did **not** define a taxonomy involving **v2**, **1B**, or **M** [2506.20380].

| Variant | Parameters | Role |
|---|---:|---|
| N | 1M | edge/on-device |
| S | 7M | low-resource |
| M | 21M | balanced default |
| L | 44M | provider-side global inference |

The family organization is central to the meaning of the name. The **“1B”** component refers to the teacher from which the student is distilled, not to the student itself. The **“M”** component denotes the medium student tier. The paper does not assign a more elaborate semantic interpretation beyond this capacity and deployment role [2607.03949].

This naming also marks a conceptual shift from TESSERA v1. The 2025 system was already framed as a **model-as-data** product based on annual global 10 m embeddings; TESSERA v2 preserves that orientation but reorganizes the system around a large teacher plus compact students designed for recurring global inference and serving [2506.20380].

## 2. Teacher and student architecture

TESSERA v2 is a **pixel-wise Earth observation foundation model** operating on **annual Sentinel-1/2 time series** at **10 m** resolution. The teacher is a **1B dual-branch pixel-wise encoder**. Each modality branch linearly embeds valid observations, adds **sinusoidal day-of-year positional encoding**, runs a **4-layer Transformer**, and aggregates the variable-length sequence by **learned attention pooling**. A **2-layer fusion Transformer** then combines the two modality tokens into a single teacher embedding
$$
t \in \mathbb{R}^{T}, \qquad T=768,
$$
followed by a final **affine-free LayerNorm**. A **batch-normalized projector** is used only during pretraining [2607.03949].

The student family, including v2-1B-M, is specified less granularly. The paper states that the four students **share the same architectural form** and differ only in **backbone width and depth**, and that each emits a **128D embedding**. Exact layer counts and widths for the 21M M student are not provided. A scaled-down version of the same pixel-wise temporal architecture is therefore a plausible implication, but the manuscript does not enumerate a full per-layer specification for v2-1B-M [2607.03949].

Architecturally, TESSERA v2 remains continuous with the first-generation system. The original TESSERA model also used modality-specific temporal encoders over **Sentinel-1 VV/VH** and **Sentinel-2 10-band MSI** annual sequences, then fused them into a **128-dimensional representation** for downstream use [2506.20380]. What changes in v2 is not the pixel-wise annual EO premise, but the scaling strategy and deployment path.

## 3. Pretraining recipe and downstream-driven scaling laws

The TESSERA v2 study fixes a **pixel-wise Barlow Twins family** and varies **encoder size**, **projector size**, and **training data** under matched compute. Teacher pretraining uses **two views per annual pixel**, with view length sampled as
$$
L \sim \mathrm{Uniform}\{8,16\},
$$
so each view is a sparse random temporal subsample of the year. Sentinel-1 and Sentinel-2 are processed in separate modality branches and then fused. The training recipe includes **Barlow Twins**, a **mix-up consistency regularizer**, **occasional whole-modality dropout**, and **global shuffling of pixels across tiles**. The reported teacher setup uses **one epoch over about 4.2 billion annual pixels**, **global batch size $131{,}072$**, **512 GPUs**, **FSDP**, and **bf16** [2607.03949].

The paper’s principal claim is that model selection by pretraining loss is ineffective for this family. Across **395 runs** evaluated on **15 downstream tasks**, converged Barlow Twins loss correlates only weakly with the composite downstream score:
$$
\text{Pearson } r = -0.18, \qquad \text{Spearman } \rho = -0.16.
$$
The reported consequence is operational rather than merely diagnostic: selecting models by loss instead of downstream score would require roughly **254\% more compute** to reach the same downstream score [2607.03949].

The resulting scaling laws are asymmetric. As pretraining compute grows, the **encoder** and the **data** should grow together, while the **projector** should remain approximately fixed:
$$
N_{\mathrm{enc}}^{\star} \propto C^{0.36}, \qquad
D^{\star} \propto C^{0.63}, \qquad
N_{\mathrm{proj}}^{\star} \propto C^{0.00}.
$$
The reported confidence intervals are $[0.29,0.45]$ for the encoder exponent, $[0.53,0.70]$ for the data exponent, and $[-0.03,0.08]$ for the projector exponent [2607.03949].

These results explain why v2-1B-M exists in its documented form. Rather than deploying a large self-supervised encoder directly, TESSERA v2 spends pretraining compute on a large teacher and matched data, keeps the projector disposable, and then distills a compact student for serving.

## 4. Distillation and Matryoshka representation structure

TESSERA v2-1B-M is not trained on task labels. It is trained by **task-agnostic representation distillation** from the frozen 1B teacher using a distillation corpus of about **200 million annual pixels**. If $s\in\mathbb{R}^{128}$ is the student embedding and $t\in\mathbb{R}^{T}$ is the frozen teacher embedding with $T=768$, then for prefix set
$$
\mathcal{K}=\{16,32,64,128\},
$$
the student objective is
$$
\mathcal{L}(s,t) \;=\; \sum_{k\in\mathcal{K}} \Bigl(1-\cos\!\bigl(h_k(s_{1:k}),\, t\bigr)\Bigr),
$$
where each temporary linear head
$$
h_k:\mathbb{R}^k \to \mathbb{R}^T
$$
maps a student prefix into the teacher space [2607.03949].

This construction defines the model’s **Matryoshka** property. The first $k$ coordinates of the 128D student vector are not trained merely to approximate the teacher’s first $k$ coordinates; instead, each prefix is trained to reconstruct the **full** 768D teacher embedding. At inference time, the embedding can therefore be truncated directly to 16, 32, or 64 coordinates as a storage-accuracy tradeoff, without retraining or auxiliary decoding [2607.03949].

The empirical retention rates reported for the student family are central to the model’s practical significance. A **16-dimensional prefix** retains about **92\%** of the full 128D composite score at **1/8 of the storage**; **32D** retains about **97\%**; and **64D** retains about **99\%** [2607.03949]. The paper further notes that truncation hurts **fine-class-count** and **regression** tasks more than **land-cover** and **change-detection** tasks, making **64D** the suggested default when task difficulty is high and storage is not the dominant constraint.

The paper also reports a negative result that motivates this design: **naive Matryoshka-style prefix losses during self-supervised pretraining fail**, because redundancy-reduction objectives identify only a subspace up to rotation. Distillation from a frozen teacher is therefore used to make prefix truncation semantically stable [2607.03949].

## 5. Benchmark performance and serving economics

The evaluation framework has two layers: a **15-task AlphaEarth suite** spanning classification, segmentation, regression, and change detection, and a **held-out 14-dataset evaluation** not used in development. Together these form a **29-task** benchmark. Downstream models are intentionally light: a **2-layer MLP** for pixel-wise tasks and a **simple CNN with fewer than 2M parameters** for patch-level tasks [2607.03949].

On the **full 29-task suite**, TESSERA v2-1B-M achieves the best reported composite score:
$$
0.611,
$$
compared with
$$
0.576 \text{ for AlphaEarth}, \quad
0.574 \text{ for TESSERA v1}, \quad
0.562 \text{ for OlmoEarth-L}.
$$
The paper explicitly notes that v2-1B-M is the **smallest** of these four compared systems [2607.03949].

On the **15-task AlphaEarth suite alone**, the ordering is slightly different. The **L** student leads at
$$
0.584,
$$
while
$$
\text{M}=0.581,\quad \text{S}=0.570,\quad \text{N}=0.558.
$$
All four exceed the directly comparable embedding products
$$
\text{AlphaEarth}=0.560,\qquad \text{TESSERA v1}=0.541.
$$
This means that v2-1B-M is the strongest **aggregate** model on the full benchmark suite, but not the strongest member of its own family on every sub-benchmark [2607.03949].

The deployment rationale is reinforced by the paper’s global inference economics. A single global annual 10 m pass with the **1B teacher** would cost about **50 H100-years**. Inference cost follows an empirical near-linear power law in encoder parameters,
$$
\mathrm{GPUy}(N) \approx 0.041\,N^{1.03},
$$
with $N$ in millions of parameters. The distilled students, by contrast, require only **0.04 to 2 H100-years** across the family [2607.03949]. The paper does not isolate a separate exact H100-year figure for the M student, but it places v2-1B-M inside that family-level range.

## 6. Embeddings-as-data deployment, continuity with TESSERA v1, and external use

TESSERA v2-1B-M is designed for **embeddings-as-data** deployment rather than end-user backbone inference. The paper states that annual global pixel embeddings are to be **precomputed and served through GeoTESSERA**, and that, upon completion of training, the authors plan to release **v2 global embeddings covering 2017–2025** [2607.03949]. This is a direct continuation of the original TESSERA program, whose principal deliverable was a set of **precomputed annual global representation maps** at **10 m** and **128 channels** for **2017–2024**, intended for frozen-encoder downstream usage and public release [2506.20380].

The v1 system already established the main EO design commitments later preserved in v2: annual **pixel-centric** Sentinel-1/Sentinel-2 representations, a **128-dimensional embedding per pixel**, and an open-source, privacy-preserving workflow in which users can consume precomputed embeddings rather than upload raw EO time series to a centralized service [2506.20380]. TESSERA v2-1B-M should therefore be understood not as a departure from that model, but as a deployment-optimized refinement of it.

External downstream evidence also points to the utility of the broader TESSERA embedding family. A later LCZ mapping study across five Swiss cities used annual **10 m**, **128-band**, **precomputed** TESSERA embeddings and reported that TESSERA outperformed both **S1S2 composites** and **AlphaEarth** in the two same-year experiments, reaching **test IoU 0.69** in the five-city setting and **test IoU 0.82** in the Bern setting [2606.20034]. That study, however, did **not** identify the checkpoint as **TESSERA v2-1B-M**; it therefore supports the TESSERA family’s downstream relevance, but not an exact per-variant attribution.

A final point of terminology follows from the publication history. The original 2025 TESSERA paper documented the model family, annual 10 m embeddings, and a large-scale implementation, but not a formal **v2-1B-M** naming scheme [2506.20380]. The exact variant identity is introduced only by the 2026 scaling paper, where v2-1B-M is defined as the **21M-parameter medium distilled student** and recommended implicitly as the practical default because it is both a **“balanced default”** and the best aggregate performer on the **29-task** suite [2607.03949].

Source: https://www.emergentmind.com/topics/tessera-v2-1b-m