© 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. 2-RES-SM
    11. 13814
    Prev
    Next
    INDEX
    Explanations

    direct or indirect expressions of emotions and interpersonal interactions

    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
    263
    +0.12
    0.7%
    102
    +0.12
    0.7%
    241
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    147
    +0.12
    0.03
    430
    +0.12
    0.03
    509
    +0.12
    0.03
    Negative Logits
    sole
    -1.59
    aterally
    -1.57
    icul
    -1.54
    ÅĽci
    -1.52
     hyperplasia
    -1.43
     defeating
    -1.43
     concomitant
    -1.43
    theless
    -1.42
    ainen
    -1.42
    rill
    -1.39
    POSITIVE LOGITS
    ©
    1.61
    ¡
    1.53
    ¬
    1.51
    watson
    1.49
    ł
    1.47
    ·
    1.41
    pole
    1.41
    ¤
    1.41
    µ
    1.40
    Ĩ
    1.39
    Activations Density 0.011%

    No Known Activations