lemur-decoding (LEMUR Decoding)

pminervini

authored 2 papers 3 months ago

Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering

Paper • 2410.15999 • Published Oct 21, 2024 • 19

Adapting Neural Link Predictors for Data-Efficient Complex Query Answering

Paper • 2301.12313 • Published Jan 29, 2023

aryopg

authored 3 papers 3 months ago

CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning

Paper • 2410.10336 • Published Oct 14, 2024 • 2

Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering

Paper • 2410.15999 • Published Oct 21, 2024 • 19

Analysing the Residual Stream of Language Models Under Knowledge Conflicts

Paper • 2410.16090 • Published Oct 21, 2024 • 7

pminervini

authored a paper 3 months ago

Attention Is All You Need But You Don't Need All Of It For Inference of Large Language Models

Paper • 2407.15516 • Published Jul 22, 2024 • 1

aryopg

authored a paper 3 months ago

DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinations

Paper • 2410.18860 • Published Oct 24, 2024 • 9

pminervini

authored a paper 3 months ago

DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinations

Paper • 2410.18860 • Published Oct 24, 2024 • 9

rohitsaxena

authored a paper 5 months ago

MovieSum: An Abstractive Summarization Dataset for Movie Screenplays

Paper • 2408.06281 • Published Aug 12, 2024 • 9

rohitsaxena

authored a paper 7 months ago

Select and Summarize: Scene Saliency for Movie Script Summarization

Paper • 2404.03561 • Published Apr 4, 2024 • 2

pminervini

authored a paper 7 months ago

A Simple and Effective $L_2$ Norm-Based Strategy for KV Cache Compression

Paper • 2406.11430 • Published Jun 17, 2024 • 23

yuzhaouoe

authored a paper 7 months ago

A Simple and Effective $L_2$ Norm-Based Strategy for KV Cache Compression

Paper • 2406.11430 • Published Jun 17, 2024 • 23

pminervini

authored 8 papers 7 months ago

PAQ: 65 Million Probably-Asked Questions and What You Can Do With Them

Paper • 2102.07033 • Published Feb 13, 2021

No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based Language Models

Paper • 2307.06440 • Published Jul 12, 2023 • 3

SPARSEFIT: Few-shot Prompting with Sparse Fine-tuning for Jointly Generating Predictions and Natural Language Explanations

Paper • 2305.13235 • Published May 22, 2023 • 1

REFER: An End-to-end Rationale Extraction Framework for Explanation Regularization

Paper • 2310.14418 • Published Oct 22, 2023 • 1

Using Natural Language Explanations to Improve Robustness of In-context Learning for Natural Language Inference

Paper • 2311.07556 • Published Nov 13, 2023

LEMUR Decoding

AI & ML interests

lemur-decoding's activity

Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering

Adapting Neural Link Predictors for Data-Efficient Complex Query Answering

CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning

Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering

Analysing the Residual Stream of Language Models Under Knowledge Conflicts

Attention Is All You Need But You Don't Need All Of It For Inference of Large Language Models

DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinations

DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinations

MovieSum: An Abstractive Summarization Dataset for Movie Screenplays

Select and Summarize: Scene Saliency for Movie Script Summarization

A Simple and Effective $L_2$ Norm-Based Strategy for KV Cache Compression

A Simple and Effective $L_2$ Norm-Based Strategy for KV Cache Compression

PAQ: 65 Million Probably-Asked Questions and What You Can Do With Them

No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based Language Models

SPARSEFIT: Few-shot Prompting with Sparse Fine-tuning for Jointly Generating Predictions and Natural Language Explanations

REFER: An End-to-end Rationale Extraction Framework for Explanation Regularization

Using Natural Language Explanations to Improve Robustness of In-context Learning for Natural Language Inference

XQA-DST: Multi-Domain and Multi-Lingual Dialogue State Tracking

Analysing The Impact of Sequence Composition on Language Model Pre-Training

An Efficient Memory-Augmented Transformer for Knowledge-Intensive NLP Tasks

AI & ML interests

Team members 5

lemur-decoding's activity