© 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. 2700
    Prev
    Next
    INDEX
    Explanations

    words related to terminology or specific concepts used in discussion

    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
    50
    +0.11
    0.4%
    1373
    +0.06
    0.2%
    1747
    +0.06
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    878
    +0.11
    0.03
    789
    +0.06
    0.03
    1586
    +0.06
    0.03
    Negative Logits
    /**
    -0.87
    
    -0.85
    <?
    -0.81
    <?
    -0.81
    <bos>
    -0.78
    ⓧ
    -0.76
    /*
    -0.75
    /***
    -0.72
    Transkript
    -0.65
     recipro
    -0.63
    POSITIVE LOGITS
     véritables
    0.74
     Kiedy
    0.66
     Dlatego
    0.65
     pymysql
    0.64
     Bardzo
    0.64
     Nasıl
    0.63
     catég
    0.63
     préoccup
    0.62
     Redacción
    0.62
     récentes
    0.61
    Activations Density 0.203%

    No Known Activations