Picture for Malvina Nissim

Malvina Nissim

QE4PE: Word-level Quality Estimation for Human Post-Editing

Add code
Mar 04, 2025
Viaarxiv icon

Multidimensional Consistency Improves Reasoning in Language Models

Add code
Mar 04, 2025
Viaarxiv icon

Are You Doubtful? Oh, It Might Be Difficult Then! Exploring the Use of Model Uncertainty for Question Difficulty Estimation

Add code
Dec 16, 2024
Figure 1 for Are You Doubtful? Oh, It Might Be Difficult Then! Exploring the Use of Model Uncertainty for Question Difficulty Estimation
Figure 2 for Are You Doubtful? Oh, It Might Be Difficult Then! Exploring the Use of Model Uncertainty for Question Difficulty Estimation
Figure 3 for Are You Doubtful? Oh, It Might Be Difficult Then! Exploring the Use of Model Uncertainty for Question Difficulty Estimation
Figure 4 for Are You Doubtful? Oh, It Might Be Difficult Then! Exploring the Use of Model Uncertainty for Question Difficulty Estimation
Viaarxiv icon

A gentle push funziona benissimo: making instructed models in Italian via contrastive activation steering

Add code
Nov 27, 2024
Viaarxiv icon

Choosy Babies Need One Coach: Inducing Mode-Seeking Behavior in BabyLlama with Reverse KL Divergence

Add code
Oct 29, 2024
Figure 1 for Choosy Babies Need One Coach: Inducing Mode-Seeking Behavior in BabyLlama with Reverse KL Divergence
Figure 2 for Choosy Babies Need One Coach: Inducing Mode-Seeking Behavior in BabyLlama with Reverse KL Divergence
Figure 3 for Choosy Babies Need One Coach: Inducing Mode-Seeking Behavior in BabyLlama with Reverse KL Divergence
Viaarxiv icon

Non Verbis, Sed Rebus: Large Language Models are Weak Solvers of Italian Rebuses

Add code
Aug 01, 2024
Viaarxiv icon

Can Model Uncertainty Function as a Proxy for Multiple-Choice Question Item Difficulty?

Add code
Jul 07, 2024
Viaarxiv icon

Multi-property Steering of Large Language Models with Dynamic Activation Composition

Add code
Jun 25, 2024
Viaarxiv icon

Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models

Add code
Jun 11, 2024
Figure 1 for Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models
Figure 2 for Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models
Figure 3 for Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models
Figure 4 for Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models
Viaarxiv icon

mCoT: Multilingual Instruction Tuning for Reasoning Consistency in Language Models

Add code
Jun 04, 2024
Viaarxiv icon