text-llm-prompted-forecast-numeric-price-series-feeding-history CLM-0222When a text LLM is prompted to forecast numeric price series, feeding the history as integer-rounded values on the original price scale gives lower forecast error than full-decimal raw values, standard-normalized values, or per-window instance-normalized values, because rounding cuts token fragmentation while keeping magnitudes in the distribution the model saw in pretraining.
Capability: Predicting future events
Observed on
Univariate hourly Bitcoin closing prices, 96-step input and 8-step horizon, 4-bit LLaMA-3 8B, zero-shot prompting for the representation ablation; single asset and single model tes.
Sources
- One-factor-at-a-time ablation on a validation set of 1441 windows with the prompt held fixed: MAPE 3.01 rounded vs 4.23 raw, 9.43 standard-normalized, 548 instance-normalized; SMAPE and NRMSE agree. Valid-output count was identical across representations, so the effect is on accuracy not format compliance. Single asset, single backbone; the tokenization explanation is argued, not measured.
Disagreeing is the most useful thing you can do here. Nothing on this page waits for a reviewer: the counts above move when evidence lands, and every source stays attached with who brought it.
Related claims
- When a text-pretrained LLM is prompted to forecast a numeric price series, rounding the input values to integers while keeping the original price scale gives lower forecast error than full-precision raw values, training-set standardization, or per-window instance normalization, because normalized values fall outside the numeric distributions the model saw in pretraining and per-window rescaling makes the same price appear differently across samples.Predicting future events · reviewed by AI
- For LLM-based forecasting systems, reported benchmark gains often cannot be attributed to temporal reasoning, because test periods predate the training cutoff, ablations show the LLM component sometimes adds nothing over the statistical or retrieval parts, and accuracy shifts under small input perturbations.Predicting future events · reviewed by AI
- Across LLM-based forecasting systems, reported accuracy gains frequently cannot be attributed to the language model's temporal reasoning, because evaluation windows predate training cutoffs, component ablations show the LLM sometimes adds nothing over the statistical or retrieval parts, and results shift under small input perturbations.Predicting future events · reviewed by AI
- On natural-language stock questions that require judging a future condition (e.g. whether a price will exceed a threshold next month), open-weight 20B-31B LLMs reasoning directly from the retrieved historical series answer near chance, and routing the question through an explicit SQL-retrieval plus time-series-forecasting toolchain raises accuracy to roughly the level of human stock investors.Predicting future events · reviewed by AI
- On numeric sequences whose first differences repeat only after a long prefix, LLaMA 3.1-8B extrapolates by computing and locally storing each first difference in its hidden states, retrieving the delta that followed the last observed one through an induction-like attention step, and adding it to the last number — so the extrapolation is a composed arithmetic operation over latent structure rather than token-level copying.Digit-level arithmetic · reviewed by AI
Notes
Drafted from the paper by a model and filed unreviewed. Visible here so it can be read, not because anyone has vouched for it: it does not move any technique's standing and does not count toward the internal scorecard. Drafted confidence: medium. Falsifier as drafted: An ablation on other assets or models where normalized or raw-decimal inputs match or beat integer-rounded inputs on scale-independent error metrics. Proposed technique, not catalogued: Integer-rounded numeric representation for LLM time-series prompts.