Model editing of a tabular search large language model using disagreement over out of distribution samples via transductive learning and contextual bandits

A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for updating a tabular search large language model (LLM), the method comprising:

• performing data pre-processing on new data associated with a tabular dataset to obtain a set of sequences,
• wherein the data pre-processing comprises:
• generating a set of sentences in a natural language each corresponding to a column of the tabular dataset and each in a predefined format,
• wherein the set of sentences are generated based on a pre-defined format;
• converting the names of each column of the tabular dataset to the natural language using the tabular search LLM;
• and
• applying prediction models to the set of sentences to generate prediction scores each associated with a pair of the set of sentences;
• and
• determining the set of sequences based on the prediction scores,
• wherein determining the set of sequences is based on whether prediction scores of a corresponding pair of one of the set of sequences meet a predefined criterion,
• wherein each column of the tabular dataset is not written in the natural language;
• applying a first fine-tuning operation on the tabular search LLM using the set of sequences,
• wherein the first fine-tuning operation comprises using a quantified low rank optimization of the set of sequences to obtain a set of sentence predictions,
• wherein the first fine-tuning operation further comprises tokenizing generated input text to obtain byte-pair encodings defined using discrete and finite vocabulary and generating follow-up tokens based on the byte-pair encodings to obtain training data comprising the byte-pair encodings and the follow-up tokens,
• wherein the training data further comprises the names of each columns and the sequences;
• applying a second fine-tuning operation on the tabular search LLM using the training data to obtain a set of final loss results and a set of updatable gradients, wherein the training data comprises at least the set of sentence predictions,
• wherein the second fine-tuning operation comprises performing a decomposition on a set of computing gradients,
• wherein the decomposition comprises:
• calculating a difference between a cumulative reward of the set of computing gradients and a sequence of actions taken by selecting a set of computing weights;
• and
• minimizing, using the difference, a regret value of the sequence of actions to obtain the set of updatable gradients;
• applying an optimization function on the set of final loss results and the set of updatable gradients to obtain optimized gradient descent parameters, wherein an updated tabular search LLM is obtained using the optimized gradient descent parameters,
• wherein the updated tabular search LLM is equipped to output information corresponding to the new data;
• and
• applying the updated tabular search LLM to a new input associated with the new data to produce a new output, wherein the new output would not be produced if the tabular search LLM is applied to the new input,
• wherein the updated tabular search LLM is not a full re-training of the tabular search LLM.

A system, comprising:

• a processor;
• and
• memory including instructions, which when executed by the processor, perform a method comprising:
• performing data pre-processing on new data associated with a tabular dataset to obtain a set of sequences,
• wherein the data pre-processing comprises:
• generating a set of sentences in a natural language each corresponding to a column of the tabular dataset and each in a predefined format,
• wherein the set of sentences are generated based on a pre-defined format;
• converting the names of each column of the tabular dataset to the natural language using the tabular search LLM;
• and
• applying prediction models to the set of sentences to generate prediction scores each associated with a pair of the set of sentences;
• and
• determining the set of sequences based on the prediction scores,
• wherein determining the set of sequences is based on whether prediction scores of a corresponding pair of one of the set of sequences meet a predefined criterion,
• wherein each column of the tabular dataset is not written in the natural language;
• applying a first fine-tuning operation on the tabular search LLM using the set of sequences,
• wherein the first fine-tuning operation comprises using a quantified low rank optimization of the set of sequences to obtain a set of sentence predictions,
• wherein the first fine-tuning operation further comprises tokenizing generated input text to obtain byte-pair encodings defined using discrete and finite vocabulary and generating follow-up tokens based on the byte-pair encodings to obtain training data comprising the byte-pair encodings and the follow-up tokens,
• wherein the training data further comprises the names of each columns and the sequences;
• applying a second fine-tuning operation on the tabular search LLM using training data to obtain a set of final loss results and a set of updatable gradients, wherein the training data comprises at least the set of sentence predictions,
• wherein the second fine-tuning operation comprises performing a decomposition on a set of computing gradients,
• wherein the decomposition comprises:
• calculating a difference between a cumulative reward of the set of computing gradients and a sequence of actions taken by selecting a set of computing weights;
• and
• minimizing, using the difference, a regret value of the sequence of actions to obtain the set of updatable gradients;
• applying an optimization function on the set of final loss results and the set of updatable gradients to obtain optimized gradient descent parameters,
• wherein an updated tabular search LLM is obtained using the optimized gradient descent parameters,
• wherein the updated tabular search LLM is equipped to output information corresponding to the new data;
• and
• applying the updated tabular search LLM to a new input associated with the new data to produce a new output, wherein the new output would not be produced if the tabular search LLM is applied to the new input,
• wherein the updated tabular search LLM is not a full re-training of the tabular search LLM.