© 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. 4-RES-SM
    11. 11913
    Prev
    Next
    INDEX
    Explanations

    references to positive aspects or advantages, commonly indicated by the word "pros."

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    307
    +0.14
    0.8%
    386
    +0.14
    0.8%
    397
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    386
    +0.14
    0.02
    194
    +0.14
    0.01
    257
    +0.14
    0.02
    Negative Logits
    ties
    -1.78
    ts
    -1.52
    ting
    -1.50
    tig
    -1.49
    KES
    -1.49
    Ĭ
    -1.48
    ±
    -1.46
    chers
    -1.46
    ¿½
    -1.45
    ski
    -1.42
    POSITIVE LOGITS
    pective
    3.15
    pero
    2.64
    per
    2.11
    odic
    1.88
    pere
    1.87
    pered
    1.87
    ily
    1.85
    icro
    1.83
    pering
    1.76
    iratory
    1.69
    Activations Density 0.213%

    No Known Activations