© 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. 24427
    Prev
    Next
    INDEX
    Explanations

    instances of "we" in various contexts

    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
    23
    +0.30
    1.8%
    369
    +0.23
    1.4%
    478
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    338
    +0.30
    0.09
    184
    +0.23
    0.08
    356
    +0.15
    0.06
    Negative Logits
    ļ
    -2.24
    ://
    -1.81
    office
    -1.68
    culus
    -1.62
    ours
    -1.59
     viewpoint
    -1.58
    powers
    -1.56
    lege
    -1.55
    ural
    -1.53
     modality
    -1.52
    POSITIVE LOGITS
    ighs
    2.02
    inberg
    1.81
    alth
    1.77
    ird
    1.73
    'll
    1.71
    eks
    1.71
    igh
    1.65
     moonlight
    1.62
    'd
    1.61
    apons
    1.61
    Activations Density 0.373%

    No Known Activations