© 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. 13188
    Prev
    Next
    INDEX
    Explanations

    references to various methods or processes

    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
    274
    +0.15
    0.9%
    287
    +0.12
    0.7%
    295
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    287
    +0.15
    0.02
    400
    +0.12
    0.01
    293
    +0.11
    0.02
    Negative Logits
    »¿
    -3.69
    ĥ½
    -3.64
    §
    -3.51
    IJ
    -3.45
    ¼
    -3.29
    ¤
    -3.18
    »
    -3.14
    Ļª
    -3.03
    ľĵ
    -3.02
    ·¸
    -3.02
    POSITIVE LOGITS
    ional
    1.90
    ologic
    1.90
    urg
    1.78
    ological
    1.77
    imetry
    1.76
    urgical
    1.73
     applied
    1.72
    imetric
    1.71
    aterial
    1.69
    utorial
    1.69
    Activations Density 0.077%

    No Known Activations