Curated · Predict Output
Top 20 Predict Output Interview Questions
Here are the top 20 AI interview questions in Predict Output format, sorted by community quality. Each entry links to a detailed answer with explanations, hints, and source-grade follow-ups. See all in this format →
Sorted by community quality · refreshed as new questions are published.
- 01Predict the output: how many tokens does this string become with a typical BPE tokenizer?Predict OutputEasy·Qual 4.0
- 02For a causal mask over a 4-token sequence, what value sits at row 2, column 3?Predict OutputEasy·Qual 4.0
- 03Estimate the token tax a single tool round-trip adds compared with an inline answerPredict OutputMedium·Qual 4.0OpenAI
- 04Predict: what is the end to end success rate of a 10-step agent task if each step succeeds with 90% probability?Predict OutputMedium·Qual 4.0
- 05Predict the break-even read count for Anthropic prompt caching given the 1.25x / 0.1x ratesPredict OutputMedium·Qual 4.0Anthropic
- 06Predict the total cost of one Claude Sonnet API call with realistic token countsPredict OutputMedium·Qual 4.0
- 07Reference: 'The cat sat on the mat.' Generated: 'A feline rested on the rug.' Predict whether BLEU-4 will be close to 0 or close to 1, and explain why.Predict OutputMedium·Qual 4.0
- 08Predict the softmax output for a causal masked attention row.Predict OutputMedium·Qual 4.0
- 09With max_grad_norm=1.0 and a raw gradient L2 of 12.5, what is the post-clip norm and scale factor?Predict OutputMedium·Qual 4.0
- 10Predict whether tokenizer.encode('world') and tokenizer.encode(' world') produce the same token IDs in tiktoken.Predict OutputMedium·Qual 4.0
- 11Predict the typical layer count for a Llama 7B and a Llama 70BPredict OutputMedium·Qual 4.0
- 12Estimate the trainable param count for a Llama-3 8B LoRA at r=16 targeting all-linear modules.Predict OutputMedium·Qual 4.0
- 13Predict supportable concurrent-user count under a 200k TPM rate limitPredict OutputMedium·Qual 4.0OpenAI
- 14Predict how many tokens the rainbow flag emoji 🏳️🌈 uses in a byte level BPE tokenizer with no emoji in training data.Predict OutputMedium·Qual 4.0
- 15Predict the shape of recall and latency on an IVF index of 10M vectors as nprobe sweeps from 1 to 50.Predict OutputMedium·Qual 4.0Pinecone
- 16The same sentence in English, Spanish, Hindi, and Japanese. Predict the token ratio on o200k_base for each.Predict OutputMedium·Qual 4.0
- 17Predict the retrieval trace: how a multi-hop RAG agent answers a chained question that single-shot RAG would fail on.Predict OutputHard·Qual 4.0
- 18Predict the throughput and per-request latency behaviour as batch size grows past saturationPredict OutputHard·Qual 4.0NVIDIA
- 19Predict the approximate token count difference between 'Hello world' and its Arabic equivalent 'مرحبا بالعالم' in a GPT-4 tokenizer.Predict OutputHard·Qual 4.0
- 20Predict the hardware cost vs gross margin for serving $15/M output tokens on H100Predict OutputHard·Qual 4.0
Related top picks
By experience level
Other formats
By topic