© 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. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 30441
    Prev
    Next
    INDEX
    Explanations

    the repeated use of the word "along"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    494
    +0.14
    0.8%
    274
    +0.13
    0.7%
    502
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    494
    +0.14
    0.02
    502
    +0.13
    0.03
    373
    +0.13
    0.02
    Negative Logits
     whistle
    -1.54
    xture
    -1.52
     Broadcasting
    -1.51
     Technology
    -1.51
     tester
    -1.45
     Breakfast
    -1.43
     permission
    -1.43
     Psychology
    -1.38
     Environment
    -1.38
     technology
    -1.38
    POSITIVE LOGITS
    µ
    2.29
    £
    2.18
    »
    2.07
    ©
    2.06
    ·
    2.06
    ł
    2.01
    ¼
    1.97
    ĨĴ
    1.95
    IJ
    1.92
    ±
    1.92
    Activations Density 0.125%

    No Known Activations