© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 0-RES-JB
    11. 326
    Prev
    Next
    INDEX
    Explanations

    the beginning of each new document

    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
    674
    +0.24
    3.6%
    1741
    +0.06
    0.8%
    1870
    +0.04
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    674
    +0.24
    0.05
    108
    +0.06
    0.24
    1713
    +0.04
    0.30
    Negative Logits
     belliger
    -1.60
     despotism
    -1.60
     unlaw
    -1.49
     ruinous
    -1.49
     massacres
    -1.43
     nukes
    -1.38
     Fascism
    -1.38
     unspeak
    -1.38
     odious
    -1.35
     demoral
    -1.34
    POSITIVE LOGITS
    <bos>
    17.44
    expandindo
    2.88
    GEBURTSDATUM
    2.87
     betweenstory
    2.84
     Administrativna
    2.73
    تقاوى
    2.70
    Autoritní
    2.69
     Italijani
    2.56
     Италијани
    2.54
     Мексичка
    2.46
    Activations Density 0.176%

    No Known Activations