© 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. 1532
    Prev
    Next
    INDEX
    Explanations

    phrases related to showcasing the best or top options in various categories or 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
    +0.13
    0.7%
    1339
    +0.05
    0.3%
    994
    +0.04
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1507
    +0.13
    0.09
    203
    +0.05
    0.08
    1339
    +0.04
    0.08
    Negative Logits
    <bos>
    -1.88
    
    -0.84
    ⓧ
    -0.77
    <?
    -0.75
    <?
    -0.73
    })();
    -0.71
    /***
    -0.70
    public
    -0.69
    ੌ
    -0.67
     mobilize
    -0.67
    POSITIVE LOGITS
     affor
    1.69
     stockholm
    1.62
     maneu
    1.62
     accla
    1.54
     increa
    1.53
     fortn
    1.52
     impra
    1.49
     véhic
    1.48
     Juf
    1.48
     Intere
    1.47
    Activations Density 0.258%

    No Known Activations