© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 472
    Prev
    Next
    INDEX
    Explanations

    financial investment terms and requirements

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    HuggingFaceFW/fineweb
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    50
    +0.21
    1.6%
    382
    +0.09
    0.6%
    2019
    +0.07
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1984
    +0.21
    0.25
    453
    +0.09
    0.23
    1034
    +0.07
    0.17
    Negative Logits
    <bos>
    -2.86
    ⓧ
    -0.99
    
    -0.90
    ///**
    -0.85
    <?
    -0.85
    /**
    -0.85
    /*
    -0.81
    //};
    -0.72
    //{
    -0.69
    public
    -0.68
    POSITIVE LOGITS
     maneu
    1.86
     increa
    1.78
     affor
    1.78
     stockholm
    1.77
     véhic
    1.72
     accla
    1.72
     impra
    1.68
     aen
    1.62
     effe
    1.60
     wien
    1.60
    Activations Density 1.677%

    No Known Activations