Fix inconsistent return type in TextEmbedder._get_sentence_embedding

[Copilot]

TextEmbedder._get_sentence_embedding returned a raw Tensor for aggregation_method='first'/'last' but a dict for 'mean', causing crashes when the caller accessed out["sentence_embedding"].

Changes:

• Return consistent dict structure {"sentence_embedding": Tensor, "label_attention_matrix": None} for all aggregation methods
• Fix indexing bug in 'last' aggregation: add .long() conversion to lengths tensor
• Update return type annotation from torch.Tensor to dict[str, Optional[torch.Tensor]]

Before:

if self.attention_config.aggregation_method == "first":
    return token_embeddings[:, 0, :]  # Raw tensor

After:

if self.attention_config.aggregation_method == "first":
    return {
        "sentence_embedding": token_embeddings[:, 0, :],
        "label_attention_matrix": None,
    }

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[Copilot]

Just as a heads up, I was blocked by some firewall rules while working on your feedback. Expand below for details.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• astral.sh
  • Triggering command: /usr/bin/curl curl -LsSf REDACTED (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)