© 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. 30681
    Prev
    Next
    INDEX
    Explanations

    references to actions or directives, specifically with the word "to."

    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
    69
    +0.22
    1.3%
    478
    +0.18
    1.0%
    23
    +0.15
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.22
    0.15
    434
    +0.18
    0.08
    295
    +0.15
    0.08
    Negative Logits
     enough
    -1.77
     among
    -1.76
    hole
    -1.66
     amongst
    -1.65
    arily
    -1.65
    ourse
    -1.62
    asio
    -1.61
    ishly
    -1.59
    sense
    -1.58
    cap
    -1.56
    POSITIVE LOGITS
    Ĥ
    3.25
    «
    3.24
    ¿½
    3.16
    ĥ½
    3.13
    Ń
    3.11
    ¥
    3.07
    ↵
    3.05
    3.05
    ↵
    3.05
    ↵↵
    3.05
    Activations Density 0.716%

    No Known Activations