© 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. 1664
    Prev
    Next
    INDEX
    Explanations

    glimpses into different aspects or areas

    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
    528
    +0.09
    0.3%
    896
    +0.07
    0.2%
    152
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    690
    +0.09
    0.03
    597
    +0.07
    0.02
    335
    +0.07
    0.02
    Negative Logits
    <?
    -0.89
    ⓧ
    -0.87
    
    -0.86
    /*
    -0.80
    /**
    -0.78
     AssemblyCompany
    -0.72
    <bos>
    -0.67
    marginVertical
    -0.66
    ||}
    -0.65
    UseVisualStyle
    -0.63
    POSITIVE LOGITS
     glimpse
    2.13
     glimpses
    1.99
     increa
    1.95
     affor
    1.86
     maneu
    1.81
     wherea
    1.81
     strick
    1.80
     stockholm
    1.78
     glimp
    1.77
     impra
    1.76
    Activations Density 0.144%

    No Known Activations