Learning From Failure: Integrating Negative Examples when Fine-tuning Large Lang

Abstract

Abstract: Large language models (LLMs) have achieved success in acting as agents, which interact with environments through tools such as search engines. However, LLMs are optimized for language generation instead of tool use during training or alignment, limiting their effectiveness as agents. To resolve this problem, previous work has first collected interaction trajectories between LLMs and environments, using only trajectories that successfully finished the task to fine-tune smaller models, making fine-tuning data scarce and acquiring it both difficult and costly. Discarding failed trajectories also

Research Question

How does the inference throughput (tokens/sec) of Llama-3-70B vary when scaling context window from 4K to 128K versus adding a 3-step retrieval pipeline on HotPotQA multi-hop questions, measured in terms of end-to-end accuracy and latency?

Verification Level

Descriptive public verification status only; aggregate claim counts are public, but individual claim records are not exposed here.

Quality Tier

Descriptive public triage only; this tier does not alter current publication or DOI behavior.

Quality Dimensions

Automated triage signals derived from public fields; not human peer review or independent validation.

Correction Record

Public corrections are additive records. Current status does not claim the synthesis is error-free.

Provenance