GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
ceselder
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
ceselder
Model Tree
Qwen/Qwen3-8B
Input Modalities
Output Modalities
Supported Functionality
Prompt (chat template, enable_thinking=False), with a single placeholder token ?:
markdown
Layer: 27?The token above contains an injected direction from this model's residual stream at the indicated layer. Write one short text snippet (roughly 30 tokens or fewer) that would make this direction activate as strongly as possible. Output only the snippet itself, with no explanation, preamble, or quotation marks.
At the placeholder position, add the norm-matched direction into the layer-1 residual stream during prefill:
python
resid[b, pos] += v_hat * resid[b, pos].norm() * 1.0 # v_hat = unit encoder column
Sample with temperature 1.0 (best-of-N over 8–64 samples strongly recommended; scoring = re-tokenize the completion standalone, run the clean base model, take max-over-positions activation).
| metric | value |
|---|---|
| greedy: mean activation vs corpus-best window (500M tokens searched) | ~0.61 |
| greedy: fraction of features beating corpus best | ~15% |
| best-of-16 @ T=1: fraction beating corpus best | ~29% |
| best-of-64 (earlier checkpoint lineage) | ~39% |
Trained/evaluated on disjoint feature splits. This checkpoint (GRPO step 10,000) maximizes held-out best-of-N performance; later checkpoints improve train reward but not transfer.
Part of a MATS interpretability project (Neel Nanda stream) on eliciting model self-knowledge of its own feature space.