replicaverse/replica-predictors

REPLICA predictor store

Per-unit predictors for computational models of human sentence processing: surprisal, several entropies, similarity-adjusted surprisal, information value and unigram surprisal, scored by language models over reading corpora (eye tracking, self-paced reading, maze, N400). Every row names the text unit it describes and the unit its number is in.

dataset	tag	predictor	stimulus_id	unit_index	text_unit	value	value_unit	layer	storage_class	unit_rule
provo	gpt2_surprisal_provo	surprisal	provo_1	4	Apple	8.4213	bits	1	positional

Start from tags.parquet: one row per (tag, predictor, file), with the model, method, configuration, storage class, layer and provenance of every cell decoded, so a cell can be found by model, method and configuration.

• 2,595 cells across 58 corpora, 285,933,322 scalar rows
• 25 methods, 26 identified models and frequency tables

Layout

path	contents
tags.parquet	the cell index: one row per (tag, predictor, file)
coverage.parquet	one row per planned or produced table with its status
history.parquet	first-publish and last-change commit of every file
<dataset>.parquet	word-grain scalar predictors of one corpus
tokens/<dataset>.parquet	token-grain scalar predictors of one corpus
characters/<dataset>.parquet	character-grain scalar predictors of one corpus
responsivity/<tag>.parquet	embedding-distance word predictors, one file per cell
lens_layers/<tag>.parquet	per-layer log-probability vectors of the lens cells, one file per cell
mcword_samples/, cont_entropy_samples/	per-unit Monte Carlo sample pools behind the entropy cells
next_log_probs/	the next-position distribution over word types at each scored word
simadj_esim/	per-unit expected-similarity arrays behind the similarity-adjusted cells
fwd_word_lookahead_entropy_samples/	look-ahead entropies at depths 1..K and the per-sample step log probabilities
manifests/<tag>.json	the compute driver's provenance sidecar of every exported cell
verification/<wave>.json	per-table verification records of one recompute wave
owt/, modelblocks_kenlm_models/, kenlm_o1/	frequency tables the unigram cells read
pool/, pool_s900000/, pool_tokens/, pool_tokens_s900000/	the marginal Monte Carlo sample pools of the pooled unigram cells

Columns

Word and character tables, and the responsivity/ files:

column	type	meaning
dataset	string	corpus, also the file name of the per-corpus tables
tag	string	cell id: model, method and configuration
predictor	string	value key, e.g. surprisal
stimulus_id	string	stimulus within the corpus
unit_index	int32	0-based position of the unit within the stimulus
text_unit	string	the unit as the corpus loader spells it; null when the stored unit count differs from the corpus
value	double	the number; NaN where undefined
value_unit	string	nats, bits, dimensionless or tokens
layer	int32	1 for REPLICA's own conventions, 2 for a study-associated computation named in the index's study column
storage_class	string	positional (stored per token, coarser units derived from the token table) or unit_bound (computed at the stored unit)
unit_rule	string	boundary rule of a unit-bound estimator; null for a positional one

Token tables (tokens/<dataset>.parquet) replace unit_index and text_unit with token_index, span_start, span_end, token_id and token. Array files add list columns whose unit is in the column's Parquet field metadata. tags.parquet carries, per cell, the model, method, configuration, realization (code commit, library versions, device, computed_at), storage columns, derivation_source for word cells aggregated from a token table, and the verification verdict with its record under verification/.

Loading

import pandas as pd

REPO = "hf://datasets/replicaverse/replica-predictors"
idx = pd.read_parquet(f"{REPO}/tags.parquet")
provo = pd.read_parquet(f"{REPO}/provo.parquet")
cell = provo[provo.tag == "gpt2_surprisal_provo"]
tokens = pd.read_parquet(f"{REPO}/tokens/provo.parquet")

Values from different models are on different scales and vocabularies; compare within a model.