### Table 8 represents the percentages related to the rejected distractors in the rst experiment together with the three di erent criteria.

"... In PAGE 6: ... Table8 : Rejected questions As it happens in the case of the accepted questions, in this case the way to compare the questions with the patterns is threefold. This time the given percentages have a di erent meaning.... ..."

### TABLE 2. THE PERCENTAGE FATFY ACID COMPOSITION OF 2 OILS COMPARED WITH THE COMPOSITION OF HUMAN SERUM CHYLOMICRONS 6 HOURS FOLLOWING INGESTION OF THE Om

### Table 1: Mean reaction times (ms) and percentage error, global F and contrasts for the three experimental conditions in each of the three tasks.

"... In PAGE 7: ...4% in the Navon task. Comparison of the three experimental conditions Mean reaction times and the proportion of trials in which participants answered a probe item incorrectly for the three experimental conditions and the three tasks are given in the first row of Table1 . For each task, a one-way analysis of variance was conducted in which the experimental condition (Target-to-Distractor, Control, Distractor-to-Target) was a within factor.... In PAGE 7: ... As reported in the second row of Table 1, these analyses showed that there were significant differences between the experimental conditions. One degree of freedom contrasts were used to compare the control condition to the Target-to-Distractor and to the Distractor-to-Target conditions (see row 3 of Table1... ..."

### Table 2: As the number of salient objects decreases and the number of distractors increases, VISOR apos;s con dence about the scene decreases gradually.

"... In PAGE 9: ...ith varying number of salient objects as well as distractor objects, i.e., those that appeared only in other scenes. Table2 (Experiment 4) shows that as the number of salient objects decreases and the number of distractors increases, VISOR found the input scene more and more confusing, and its con dence about the scene was lowered gradually. The activity of the workbench schema was a ected by the competition from other scene schemas.... In PAGE 10: ... As in the previous experiment, VISOR was also presented test scenes that vary in the number of salient objects and distractors for recognition. This time, its con dence about the scenes decreased less ( Table2 , Experiment 5) compared to the previous experiment (Table 2, Experiment 4) because the missing salient objects appeared only 50% of the time during training. Therefore, VISOR uses the encoded probabilities to judge how important an object identi es a scene, which further adds to its exibility in representing \soft quot; schemas.... ..."

### Table 3: Task 2c: LSTM with very long minimal time lags q + 1 and a lot of noise. p is the number of available distractor symbols (p + 4 is the number of input units). q p is the expected number of occurrences of a given distractor symbol in a sequence. The rightmost column lists the number of training sequences required by LSTM (BPTT, RTRL and the other competitors have no chance of solving this task). If we let the number of distractor symbols (and weights) increase in proportion to the time lag, learning time increases very slowly. The lower block illustrates the expected slow-down due to increased frequency of distractor symbols.

"... In PAGE 15: ... 20 trials were made for all tested pairs (p; q). Table3 lists the mean of the number of training sequences required by LSTM to achieve success (BPTT and RTRL have no chance of solving non-trivial tasks with minimal time lags of 1000 steps). Scaling.... In PAGE 15: ... Scaling. Table3 shows that if we let the number of input symbols (and weights) increase in proportion to the time lag, learning time increases very slowly. This is a another remarkable property of LSTM not shared by any other method we are aware of.... In PAGE 15: ... Distractor in uence. In Table3 , the column headed by q p gives the expected frequency of distractor symbols. Increasing this frequency decreases learning speed, an e ect due to weight oscillations caused by frequently observed input symbols.... ..."

### TABLE 2. PROBE RTS (IN MS) FOR OFF-PROBES AND FOR ON-DISTRACTOR-PROBES WITH EITHER A CHANGE OR NO- CHANGE TO THE DISTRACTOR PRIOR TO PROBE PRESENTATION. ALSO PRESENTED ARE THE CORRESPONDING ON- PROBE RT COSTS.

### Table IV. Comparison between configurations of the cache model for pronouns Repl Policy Cache Size ACC AmbAve Distractors

2006

Cited by 1

### Table 2. * marks the most frequent (initial) prefixes. Other (infrequent) prefixes merely fill the gaps

1997

"... In PAGE 8: ... Note that ad is an edge. All frequent and newly added, infrequent prefixes are summarized in Table2 . Note that in this example, only one infrequent edge ad is produced to fill the gap between the two frequent peaks ac and adbb .... ..."

Cited by 7

### Table 3: Task 2c: LSTM with very long minimal time lags q+1 and a lot of noise. p is the number of available distractor symbols (p + 4 is the number of input units). q p is the expected number of occurrences of a given distractor symbol in a sequence. The last column lists the number of training sequences required by LSTM (of course, BPTT/RTRL and the other competitors have no chance of solving non-trivial tasks with minimal time lags involving 1000 time steps). If we let the number of distractor symbols (and weights) increase in proportion to the time lag, learning time increases very slowly. The lower block illustrates the expected slow-down due to increased frequency of distractor symbols.

"... In PAGE 13: ...equences.Results. 20 trials were made for all tested pairs (p; q). Table3 lists the mean of the number of training sequences required by LSTM to achieve success (of course, BPTT and RTRL have no chance of solving non-trivial tasks with minimal time lags involving 1000 time steps). Scaling.... In PAGE 13: ... Scaling. Table3 shows: if we let the number of input symbols (and weights) increase in proportion to the time lag, learning time increases very slowly. This is a another remarkable property of LSTM not shared by any other method we are aware of.... In PAGE 13: ... Distractor in uence. In Table3 , the column headed by q p re ects the expected frequency of distractor symbols. Increasing this frequency decreases learning speed.... ..."

### Table 1 Experiment 1: Mean Response Latencies (in Milliseconds), varied by Distractor Type (Visual vs. Auditory), Relatedness (Unrelated vs. Semantic vs. Phonological) and Picture-Word Onset Asynchrony (SOA)

"... In PAGE 4: ....0%, and 1.4% of the data, respectively. A total of 3.4% of data points were deleted in this way. Table1 lists the mean response latencies, varied by Dis- tractor Modality, Target-Distractor Relation, and SOA. Fig- ure 1 displays the semantic and phonological effects (Unre- lated minus Related condition).... ..."