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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 2400
    Prev
    Next
    INDEX
    Explanations

    real estate listings emphasizing properties with large features

    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
    +0.12
    0.6%
    1265
    +0.07
    0.3%
    1145
    +0.07
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    523
    +0.12
    0.06
    369
    +0.07
    0.06
    1084
    +0.07
    0.06
    Negative Logits
    <bos>
    -1.88
    ⓧ
    -0.96
    <?
    -0.81
    
    -0.76
    /**
    -0.75
    <?
    -0.73
    ///**
    -0.71
    ുറ
    -0.71
    setClass
    -0.66
    //
    -0.65
    POSITIVE LOGITS
     maneu
    1.64
     stockholm
    1.60
     hcm
    1.52
     wien
    1.48
     lara
    1.48
     emphat
    1.47
     LARGE
    1.44
     large
    1.44
     Large
    1.43
     larges
    1.43
    Activations Density 0.113%

    No Known Activations