© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 28334
    Prev
    Next
    INDEX
    Explanations

    phrases indicating partial or substantial influence or impact

    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
    288
    +0.13
    0.7%
    276
    +0.12
    0.7%
    59
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.03
    500
    +0.12
    0.03
    102
    +0.11
    0.02
    Negative Logits
    ologie
    -1.60
    xtures
    -1.59
    "),
    -1.56
    biamo
    -1.47
    '),
    -1.45
    ”),
    -1.44
    ots
    -1.44
    ipeg
    -1.42
     website
    -1.41
     fa
    -1.37
    POSITIVE LOGITS
     unchanged
    1.76
    cular
    1.70
    ulence
    1.67
    ifies
    1.59
     unaffected
    1.57
    ulent
    1.54
     obscured
    1.51
    isans
    1.48
     overlooked
    1.47
     leng
    1.47
    Activations Density 0.139%

    No Known Activations