© 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. 1-RES-SM
    11. 27967
    Prev
    Next
    INDEX
    Explanations

    instances of the word "appointed."

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    156
    +0.17
    0.9%
    287
    +0.11
    0.6%
    148
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    251
    +0.17
    0.03
    287
    +0.11
    0.03
    150
    +0.11
    0.03
    Negative Logits
    Ĵ
    -2.13
    Īĺ
    -1.98
    ĵ
    -1.66
    )$-
    -1.64
    IJ
    -1.63
    uality
    -1.63
    iness
    -1.47
    Ļ
    -1.46
    ı
    -1.42
     similarities
    -1.42
    POSITIVE LOGITS
     appointment
    1.88
     appointments
    1.70
     mistrial
    1.67
     pardon
    1.63
     appointed
    1.53
     chairman
    1.52
     consultation
    1.51
     chair
    1.51
     counsel
    1.50
     guide
    1.50
    Activations Density 0.245%

    No Known Activations