Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Initial Gemma-2B SAEs
    3. Gemma-2B
    4. Residual Stream
    5. 0-RES-JB
    6. 182
    Prev
    Next
    INDEX
    Explanations

    phrases related to political discussion and technical jargon used in various domains

    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
    +0.20
    0.8%
    2019
    +0.09
    0.4%
    1265
    +0.07
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1265
    +0.20
    0.07
    382
    +0.09
    0.07
    2019
    +0.07
    0.07
    Negative Logits
    <bos>
    -2.29
    ⓧ
    -0.97
    <?
    -0.85
    /**
    -0.82
    
    
    -0.82
    /***
    
    -0.80
    /*
    -0.80
    Vegeu
    -0.79
    /*++
    -0.75
     "..\..\..\
    -0.75
    POSITIVE LOGITS
     affor
    2.05
     impra
    1.96
     maneu
    1.95
     accla
    1.94
     increa
    1.77
     unden
    1.74
     shenan
    1.72
     unspeak
    1.72
     stockholm
    1.71
     perfet
    1.69
    Activations Density 0.774%

    No Known Activations