© 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. 12620
    Prev
    Next
    INDEX
    Explanations

    events and information related to occurrences or activities

    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
    156
    +0.17
    1.0%
    258
    +0.12
    0.7%
    219
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    38
    +0.17
    0.02
    481
    +0.12
    0.02
    299
    +0.11
    0.02
    Negative Logits
    ¸
    -3.20
    ĥ½
    -2.82
    IJ
    -2.80
    İ
    -2.78
    ¿½
    -2.74
    ·
    -2.71
    ³
    -2.65
    ŀ
    -2.53
    į
    -2.52
    ↵Č
    -2.50
    POSITIVE LOGITS
    uality
    3.23
    uate
    2.26
    ually
    2.23
    uating
    2.15
    uelle
    2.04
    uated
    1.83
    ional
    1.82
    uates
    1.81
    sworth
    1.81
    uary
    1.80
    Activations Density 0.213%

    No Known Activations