© 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. 1607
    Prev
    Next
    INDEX
    Explanations

    mentions of military operations and attacks

    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
    1241
    +0.06
    0.2%
    235
    +0.06
    0.1%
    1594
    +0.06
    0.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    283
    +0.06
    0.02
    1343
    +0.06
    0.03
    1871
    +0.06
    0.01
    Negative Logits
     kram
    -1.36
     utop
    -1.22
     lele
    -1.20
     labd
    -1.20
     thuy
    -1.18
     plak
    -1.17
     bont
    -1.16
    :))
    -1.16
     blin
    -1.16
     franz
    -1.14
    POSITIVE LOGITS
    ¬
    1.93
     actionTypes
    0.75
     ¬
    0.73
     smtplib
    0.67
    comuna
    0.65
    pośred
    0.61
     bénéfice
    0.60
     églises
    0.60
    municipi
    0.59
     wonen
    0.59
    Activations Density 0.222%

    No Known Activations