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. 72
    Prev
    Next
    INDEX
    Explanations

    technical terms related to optimization and operation management, with a focus on balancing aspects in various contexts

    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
    +0.07
    0.3%
    1262
    +0.06
    0.2%
    420
    +0.06
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1706
    +0.07
    0.03
    2009
    +0.06
    0.04
    1660
    +0.06
    0.03
    Negative Logits
    <bos>
    -1.18
    /***
    
    -0.86
    public
    -0.66
    ///**
    -0.60
    qian
    -0.59
    private
    -0.58
    /**
    
    -0.58
    displayquote
    -0.57
    /**
    -0.56
    else
    -0.56
    POSITIVE LOGITS
     balancing
    1.72
     Balancing
    1.60
     balanced
    1.49
    Balancing
    1.41
    balancing
    1.41
     unbalanced
    1.40
     balances
    1.38
     Balance
    1.37
     Balanced
    1.36
     balance
    1.35
    Activations Density 0.235%

    No Known Activations