© 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. 14720
    Prev
    Next
    INDEX
    Explanations

    past and present tense forms of the verb "to be."

    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
    23
    +0.15
    0.9%
    18
    +0.12
    0.7%
    325
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.15
    0.02
    53
    +0.12
    0.02
    214
    +0.10
    0.02
    Negative Logits
    ↵↵
    -2.71
    ↵
    -2.71
    ↵
    -2.71
    -2.71
    ↵
    -2.71
    ↵
    -2.71
    ↵
    -2.71
    č↵
    -2.71
    -2.71
    č↵č↵
    -2.71
    POSITIVE LOGITS
     born
    2.09
     president
    1.94
     Professor
    1.92
     chairman
    1.84
     king
    1.79
     nephew
    1.71
     proud
    1.70
     eldest
    1.69
     Deputy
    1.69
     professor
    1.68
    Activations Density 0.034%

    No Known Activations