© 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. 14497
    Prev
    Next
    INDEX
    Explanations

    commands and configurations related to Docker usage

    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
    411
    +0.15
    0.8%
    29
    +0.13
    0.7%
    69
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.15
    0.04
    405
    +0.13
    0.04
    157
    +0.11
    0.06
    Negative Logits
     acupuncture
    -2.00
     pupils
    -1.87
     Sax
    -1.73
     deaf
    -1.69
     acup
    -1.69
     teachers
    -1.67
     pupil
    -1.66
     textbooks
    -1.66
     Ala
    -1.62
    student
    -1.58
    POSITIVE LOGITS
     Docker
    4.46
     docker
    4.24
    docker
    3.74
     container
    3.19
     containers
    3.19
     Container
    3.13
     Kubernetes
    3.03
    container
    2.58
    ĻĤ
    2.48
    yml
    2.38
    Activations Density 1.156%

    No Known Activations