© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 2808
    Prev
    Next
    INDEX
    Explanations

    references to religious themes, particularly related to concepts of grace and mercy

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    650
    +0.09
    0.3%
    795
    +0.08
    0.3%
    1065
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1094
    +0.09
    0.02
    795
    +0.08
    0.02
    1776
    +0.07
    0.02
    Negative Logits
    public
    -0.83
    ‏‏
    -0.80
    int
    -0.80
    #
    -0.77
    BUYER
    -0.75
    static
    -0.74
    return
    -0.74
    uint
    -0.74
    //
    -0.73
    protected
    -0.73
    POSITIVE LOGITS
     grace
    2.60
     maneu
    2.37
     GRACE
    2.32
     reluct
    2.22
    grace
    2.21
     shenan
    2.17
     accla
    2.16
     Grace
    2.15
     increa
    2.14
     disreg
    2.12
    Activations Density 0.126%

    No Known Activations