© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 1054
    Prev
    Next
    INDEX
    Explanations

    phrases related to structured objects like lines or rows

    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%
    1437
    +0.06
    0.2%
    1614
    +0.05
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1082
    +0.07
    0.03
    1052
    +0.06
    0.03
    1494
    +0.05
    0.03
    Negative Logits
    <bos>
    -1.26
    ⓧ
    -0.95
    /**
    -0.89
    /***
    -0.78
    
    -0.71
    Fordítás
    -0.65
    <?
    -0.65
    public
    -0.64
    <?
    -0.63
     prochaines
    -0.62
    POSITIVE LOGITS
     lines
    1.95
     Lines
    1.79
     LINES
    1.66
    Lines
    1.50
    lines
    1.44
     jaya
    1.36
     haup
    1.31
     hcm
    1.25
     Somal
    1.24
     bandung
    1.24
    Activations Density 0.187%

    No Known Activations