© 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. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 13813
    Prev
    Next
    INDEX
    Explanations

    references to medical studies and their citation formats

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    21
    +0.21
    1.2%
    216
    +0.13
    0.8%
    282
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    68
    +0.21
    0.02
    216
    +0.13
    0.02
    21
    +0.11
    0.01
    Negative Logits
     clarity
    -1.62
     majority
    -1.52
    Inflater
    -1.50
     stranger
    -1.50
     edited
    -1.48
     coverage
    -1.41
     distribution
    -1.40
    izations
    -1.37
     course
    -1.36
    isation
    -1.35
    POSITIVE LOGITS
    į
    3.70
    ĸ
    3.61
    ´
    3.59
    ¤
    3.58
    IJ
    3.54
    °
    3.46
    »
    3.45
    ı
    3.42
    Ķ
    3.38
    ·
    3.35
    Activations Density 0.039%

    No Known Activations