© 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. 30698
    Prev
    Next
    INDEX
    Explanations

    copyright-related issues and violations

    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
    218
    +0.16
    0.9%
    469
    +0.12
    0.7%
    308
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    218
    +0.16
    0.04
    469
    +0.12
    0.04
    97
    +0.11
    0.03
    Negative Logits
    HECK
    -1.66
    upon
    -1.58
    öt
    -1.58
    ower
    -1.54
    ligt
    -1.50
    acters
    -1.45
    inae
    -1.42
    ait
    -1.42
    amycin
    -1.42
    itrile
    -1.41
    POSITIVE LOGITS
    ¦
    1.98
    ĻĤ
    1.94
    ¬
    1.87
    ¾
    1.87
    Ŀ
    1.85
    ¡
    1.70
    rated
    1.66
    ĸ
    1.61
     news
    1.61
    ķ
    1.61
    Activations Density 0.175%

    No Known Activations