© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 15091
    Prev
    Next
    INDEX
    Explanations

    references to employment and job-related activities

    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
    +0.11
    0.6%
    473
    +0.11
    0.6%
    417
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    318
    +0.11
    0.05
    398
    +0.11
    0.02
    184
    +0.11
    0.05
    Negative Logits
    apine
    -1.69
    Ĩ
    -1.63
    Ibid
    -1.63
    Ħ
    -1.60
    ications
    -1.57
    cticut
    -1.52
    inctions
    -1.44
    abulary
    -1.41
    riber
    -1.39
    ľ
    -1.39
    POSITIVE LOGITS
     dilig
    2.43
     alongside
    2.34
     overtime
    2.28
    hops
    2.27
     collabor
    2.27
    paces
    2.25
     together
    2.23
    hop
    2.17
     tire
    2.17
     closely
    2.08
    Activations Density 0.268%

    No Known Activations