© 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. 2-RES-SM
    11. 32236
    Prev
    Next
    INDEX
    Explanations

    the term "ich" related to a specific cultural or contextual theme

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    505
    +0.12
    0.7%
    391
    +0.12
    0.7%
    444
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    391
    +0.12
    0.02
    59
    +0.12
    0.01
    13
    +0.12
    0.01
    Negative Logits
    ção
    -1.60
    herent
    -1.55
    epid
    -1.45
    eners
    -1.44
    xit
    -1.42
    ening
    -1.41
    ttes
    -1.40
    escence
    -1.40
    irectional
    -1.40
    epend
    -1.39
    POSITIVE LOGITS
    olson
    1.95
    igan
    1.86
    keit
    1.74
    olas
    1.69
    lain
    1.63
    schild
    1.62
    annel
    1.59
    uman
    1.56
    oli
    1.56
    ael
    1.55
    Activations Density 0.117%

    No Known Activations