Gemma 2-2B
PulseAugur coverage of Gemma 2-2B — every cluster mentioning Gemma 2-2B across labs, papers, and developer communities, ranked by signal.
- instance of Gemma 2-9B 90%
- used by Qwen2.5-1.5B 70%
- used by Sparse Autoencoders 70%
- used by alphaXiv 70%
- used by Gotit.pub 70%
- used by DagsHub 70%
- instance of Gemma 3:1B 70%
- instance of CatalyzeX 70%
- developed Sparse Autoencoders 70%
- used by arXiv 60%
- competes with Llama 3.2:3b 60%
- competes with Llama-3.2:1b 60%
8 day(s) with sentiment data
-
New benchmark probes LLM confabulation vs. refusal in hidden states
Researchers have developed a new benchmark called ConfabQA to analyze how small language models handle unanswerable questions. The study investigates whether the model's internal state before generating a response can p…
-
New research explores efficient test-time reasoning for LLMs · 10 sources tracked
Multiple research papers explore novel methods for enhancing the reasoning capabilities of large language models (LLMs) at test time. Techniques include learning when to apply teacher supervision to prevent "Teacher Dri…
-
New CHaRS method steers language models to reduce toxicity
Researchers have developed a new method for steering language model activations, called Clustered Optimal-Transport Steering (CHaRS). This technique uses optimal transport to pair clusters of classes and moves hidden st…
-
New method improves entity-replacement responses in LLMs like Qwen, Llama
Researchers have explored a method called hidden-state reuse to improve how language models like Qwen, Llama, and Gemma-2-2B handle entity-replacement responses. This technique involves injecting changes into the model'…
-
Research probes emotion geometry in Google's Gemma 2 LLMs
A new research paper explores the geometric representation of emotions within large language models, specifically examining Google's Gemma 2 family. The study found that all six Gemma 2 models analyzed, across various s…
-
New TopKLoRA method enables interpretable AI model fine-tuning and backdoor removal
Researchers have developed TopKLoRA, an interpretable fine-tuning method designed to identify and remove malicious backdoors in AI models. This technique, an extension of prior work, activates only the largest adapter l…
-
Local AI tool processes notes using Gemma 2-2B
A new tool utilizes the Gemma 2-2B model to process a small number of notes, limiting its operation to five notes and a two-minute duration. This tool is designed to run locally on a laptop CPU and will not process note…
-
AI Research Explores Reasoning, Legal Concepts, and Model Biases
Recent research explores various facets of AI, including improvements to recursive reasoning models with Gated Answer Resampling (GAR) that significantly boost performance on complex puzzles. Another study investigates …
-
New research explores Sparse Autoencoders for LLM interpretability and control · 5 sources tracked
Researchers are exploring Sparse Autoencoders (SAEs) to better understand and control the internal workings of large language models. New methods like the Redistribution Sensitivity Score (RSS) and SAE Additivity Residu…
-
New Locket Framework Enhances LLM Privacy and Access Control
Researchers have developed a new framework called Locket that enhances security and privacy for large language models (LLMs). Locket uses lightweight adapters, each associated with a specific access policy, and a gating…
-
Gemma-2-2B outperforms larger LLMs on narrative infilling benchmark
A new benchmark for narrative infilling, designed to evaluate how well Large Language Models can reconstruct missing sentences in stories, has been introduced. The benchmark, comprising approximately 9.2K instances acro…
-
Small LLMs achieve 10/10 task success on old phones with new SiFR layer
A new layer called SiFR (Systemic Information Filtering and Retrieval) enables small language models to perform complex tasks on older mobile devices. Researchers demonstrated that a 400MB Qwen3-0.6B model running on a …
-
New research tackles SAE limitations and explores hierarchical structures
Recent research explores the limitations and advancements in sparse autoencoders (SAEs) and related techniques for understanding and processing complex data. One paper investigates why SAEs struggle with compositional g…
-
New LLMs Developed for Moroccan Arabic and Arabizi Dialects
A developer has fine-tuned two open-source large language models, SILMA-9B-Darija and SILMA-2B-Darija, to better understand and generate Moroccan Arabic (Darija) and its informal Arabizi script. These models were traine…
-
New framework offers formal guarantees for LLM interpretability
A new formal verification framework has been developed to address the fragility of mechanistic interpretability in large language models. Researchers demonstrated that minor input changes can drastically alter the inter…
-
New AI alignment methods improve efficiency and multi-dimensional control · 3 sources tracked
Researchers are developing new methods for aligning AI models with human preferences, aiming to improve efficiency and performance. One approach, DSPA, uses inference-time steering to condition alignment on prompts, sho…
-
Small Qwen3 LLM on Old Phone Controls Desktop Browser
A demonstration showcases the Qwen3-0.6B language model, running on a 2017 Samsung Note 8, successfully controlling a desktop Google Chrome browser. The model processed structured page representations to perform tasks l…
-
Cross-model KV cache sharing promises to speed up multi-model AI inference
Two research papers propose a method called cross-model KV cache sharing to improve the efficiency of multi-model AI inference pipelines. This technique allows the key-value states computed by one model during its initi…
-
New 'patterning' technique debiases AI reward models, shows cross-model transfer
Researchers have developed a new technique called "patterning" to debias reward models used in AI training. This method reweights preference pairs based on their impact on benchmark losses, effectively reducing stylisti…
-
Hugging Face unveils efficient multimodal encoder NeoMME, study favors encoders for Indic NER
Hugging Face has introduced NeoMME, a new family of multilingual multimodal encoders designed for efficiency. Unlike many generative models, NeoMME uses a single bidirectional Transformer to process both text and image …