token-level

Tag

Cards List
#token-level

Token-Level Off-Policy Learning for Faithful Generation Under Distribution Shift

Hugging Face Daily Papers · yesterday Cached

Proposes Token-Level Off-Policy Labeling (TOPL), an off-policy training paradigm for faithful generation that reframes post-training as token-level correctness prediction, achieving strong out-of-distribution generalization across summarization and machine translation tasks.

0 favorites 0 likes
#token-level

OriginBlame: Record- and Token-Level Data Provenance for AI Training Datasets

arXiv cs.AI · 5d ago Cached

OriginBlame is a record- and token-level data provenance system that propagates author identity through AI training data pipelines, enabling precise forget sets for machine unlearning. It eliminates over-deletion from dataset-level systems and improves unlearning effectiveness.

0 favorites 0 likes
#token-level

Consensus as Privileged Context for Label-Free Self-Distillation

arXiv cs.LG · 5d ago Cached

A research paper introducing Canon, a label-free self-distillation method that uses consensus among sampled solutions to provide dense token-level supervision for training large language models on reasoning tasks, improving pass@1 by up to 12 points and outperforming label-free reinforcement learning at a fraction of the compute.

0 favorites 0 likes
#token-level

When Implausible Tokens Get Reinforced: Tail-Aware Credit Calibration for LLM Reinforcement Learning

arXiv cs.CL · 2026-07-10 Cached

Proposes Tail-Aware Credit Calibration (TACO) to address positive-credit contamination in LLM reinforcement learning by calibrating uniform credit assignment to suppress updates on implausible tail tokens, improving training stability and performance.

0 favorites 0 likes
#token-level

Attending to Multimodal Generation One Token at a Time

Hugging Face Daily Papers · 2026-07-04 Cached

This paper investigates token-level attention shifts in multimodal large language models during generation, revealing consistent patterns and proposing a simple test-time intervention that significantly improves task performance.

0 favorites 0 likes
#token-level

TokenScope: Token-Level Explainability and Interpretability for Code-Oriented Tasks in Large Language Models

arXiv cs.CL · 2026-07-03 Cached

TokenScope is an interactive interpretability tool for decoder-only large language models that provides token-level metrics, attention patterns, and counterfactual branching during code generation, enabling systematic investigation of model behavior.

0 favorites 0 likes
#token-level

CORTEX: Token-Level Hallucination Detection in RAG via Comparative Internal Representations

arXiv cs.CL · 2026-07-01 Cached

Proposes CORTEX, a token-level hallucination detection method for RAG that compares LLM internal representations with and without retrieved documents to identify ungrounded spans. It improves fine-grained localization of hallucinations in long-form RAG outputs.

0 favorites 0 likes
#token-level

Comparing Transformers and Hybrid Models at the Token Level

Lobsters Hottest · 2026-06-27 Cached

This paper analyzes token-level prediction differences between transformers and hybrid attention-recurrent models using Olmo 3 and Olmo Hybrid, finding that hybrids improve on semantic state tracking while transformers excel at n-gram copying and syntactic bracket matching.

0 favorites 0 likes
#token-level

Beyond Entropy: Learning from Token-Level Distributional Deviations for LLM Reasoning

arXiv cs.AI · 2026-06-20 Cached

Introduces Independent Combinatorial Tokens (ICT) framework that uses Jensen-Shannon divergence between token logit distributions to identify critical branching points, preventing entropy collapse and explosion in RLVR for LLM reasoning. Achieves up to 14.9% pass@4 improvement on Qwen models.

0 favorites 0 likes
#token-level

STARE: Surprisal-Guided Token-Level Advantage Reweighting for Policy Entropy Stability

Hugging Face Daily Papers · 2026-06-17 Cached

STARE addresses policy entropy collapse in GRPO-based reinforcement learning for large language models by introducing surprisal-guided token-level advantage reweighting and target-entropy regulation, achieving 4%-8% accuracy gains on AIME benchmarks.

0 favorites 0 likes
#token-level

How Language Models Fail: Token-Level Signatures of Committed and Persistent Reasoning Failures

arXiv cs.CL · 2026-06-08 Cached

This paper characterizes two distinct processes by which language models fail in reasoning—committed failure and persistent uncertainty—using token-level uncertainty signals, and demonstrates implications for self-consistency and failure detection strategies.

0 favorites 0 likes
#token-level

ARCA: Adapter-Residual Credit Assignment When Token Signals Degenerate

arXiv cs.LG · 2026-06-02 Cached

This paper identifies a structural failure mode in token-level credit assignment for LLM reinforcement learning when using LoRA, where intrinsic signals degenerate. It proposes Adapter-Residual Credit Assignment (ARCA), which derives token salience from adapter hidden-state residuals and remains competitive with baselines.

0 favorites 0 likes
#token-level

RAGognizer: Hallucination-Aware Fine-Tuning via Detection Head Integration

arXiv cs.CL · 2026-04-20 Cached

RAGognizer introduces a hallucination-aware fine-tuning approach that integrates a lightweight detection head into LLMs for joint optimization of language modeling and hallucination detection in RAG systems. The paper presents RAGognize, a dataset of naturally occurring closed-domain hallucinations with token-level annotations, and demonstrates state-of-the-art hallucination detection while reducing hallucination rates without degrading language quality.

0 favorites 0 likes
← Back to home

Submit Feedback