REVIEW 4 cited by
Positional Description Matters for Transformers Arithmetic
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Transformers, central to the successes in modern Natural Language Processing, often falter on arithmetic tasks despite their vast capabilities --which paradoxically include remarkable coding abilities. We observe that a crucial challenge is their naive reliance on positional information to solve arithmetic problems with a small number of digits, leading to poor performance on larger numbers. Herein, we delve deeper into the role of positional encoding, and propose several ways to fix the issue, either by modifying the positional encoding directly, or by modifying the representation of the arithmetic task to leverage standard positional encoding differently. We investigate the value of these modifications for three tasks: (i) classical multiplication, (ii) length extrapolation in addition, and (iii) addition in natural language context. For (i) we train a small model on a small dataset (100M parameters and 300k samples) with remarkable aptitude in (direct, no scratchpad) 15 digits multiplication and essentially perfect up to 12 digits, while usual training in this context would give a model failing at 4 digits multiplication. In the experiments on addition, we use a mere 120k samples to demonstrate: for (ii) extrapolation from 10 digits to testing on 12 digits numbers while usual training would have no extrapolation, and for (iii) almost perfect accuracy up to 5 digits while usual training would be correct only up to 3 digits (which is essentially memorization with a training set of 120k samples).
Forward citations
Cited by 4 Pith papers
-
DriveCode: Domain Specific Numerical Encoding for LLM-Based Autonomous Driving
DriveCode's continuous number projector and regression number head reduce control-signal errors in LLM autonomous driving compared with text-token and xVal baselines on the tested datasets.
-
Explaining and Tuning Transformer-based LLMs in Arithmetic Tasks with Human Strategies
A small transformer learns addition, multiplication, and division by mastering simple digit subtasks first, and human teaching strategies lift its arithmetic accuracy to ~100%.
-
Learning Moderately Input-Sensitive Functions: A Case Study in QR Code Decoding
A Transformer can learn to decode QR codes from bit strings, outperforming standard error correction on corrupted codes by exploiting language structure, but only under a simplified fixed-mask setup.
-
Extrapolation by Association: Length Generalization Transfer in Transformers
Length generalization on a short-trained main task can be inherited from a longer-trained related auxiliary task trained jointly with it.
Discussion (0). Sign in to comment.