© 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. 10-RES-JB
    11. 1373
    Prev
    Next
    INDEX
    Explanations

    instances of the word "dispute" and related contexts indicating conflict or contention

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.10.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    chanind/openwebtext-gemma
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    standard
    Context Size
    1,024
    Dataset
    ctigges/openwebtext-gemma-1024-cl
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    50
    +0.20
    1.1%
    411
    +0.16
    0.9%
    1325
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    411
    +0.20
    0.02
    1325
    +0.16
    0.02
    889
    +0.15
    0.02
    Negative Logits
    <bos>
    -2.84
    add
    -0.66
     realize
    -0.63
     add
    -0.62
    von
    -0.61
    /*++
    -0.60
     let
    -0.60
    <?
    -0.60
    char
    -0.59
    contentLoaded
    -0.59
    POSITIVE LOGITS
     milf
    1.74
     disagre
    1.65
     🤣🤣
    1.64
     unspeak
    1.56
     😭😭
    1.54
     affor
    1.52
     maneu
    1.52
     indescri
    1.50
     madonna
    1.47
     hentai
    1.47
    Activations Density 0.040%

    No Known Activations