© 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. 4-RES-SM
    11. 10810
    Prev
    Next
    INDEX
    Explanations

    references to legal rules and regulations

    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
    78
    +0.12
    0.7%
    468
    +0.11
    0.6%
    236
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    168
    +0.12
    0.04
    157
    +0.11
    0.04
    432
    +0.11
    0.03
    Negative Logits
     impression
    -1.82
    ity
    -1.68
    aju
    -1.63
    stage
    -1.60
    adays
    -1.51
    ry
    -1.49
    onson
    -1.49
    /%
    -1.47
     reality
    -1.46
    ance
    -1.45
    POSITIVE LOGITS
    Į
    3.33
    İ
    3.29
    Ĥ
    3.25
    Ī
    3.16
    Ļ
    3.15
    ĻĤ
    3.14
    Ķ
    3.05
    IJ
    3.05
    °
    3.04
    ŀ
    3.02
    Activations Density 0.226%

    No Known Activations