© 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. 1-RES-SM
    11. 28994
    Prev
    Next
    INDEX
    Explanations

    terms related to direct and indirect actions or conversions

    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
    404
    +0.14
    0.8%
    438
    +0.13
    0.7%
    275
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    404
    +0.14
    0.03
    438
    +0.13
    0.02
    391
    +0.11
    0.02
    Negative Logits
     Chiefs
    -1.63
    aves
    -1.61
    mary
    -1.60
    cules
    -1.55
    pace
    -1.45
    yla
    -1.43
     new
    -1.39
     Winnipeg
    -1.39
    odot
    -1.39
    days
    -1.37
    POSITIVE LOGITS
    amente
    2.22
    ivity
    1.90
    itud
    1.87
    erad
    1.75
     Redistributions
    1.59
    mente
    1.57
    iento
    1.55
    going
    1.52
    /,
    1.50
    /
    1.49
    Activations Density 0.068%

    No Known Activations