© 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. 27486
    Prev
    Next
    INDEX
    Explanations

    various types of opening parenthesis characters

    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
    478
    +0.24
    1.4%
    219
    +0.18
    1.0%
    369
    +0.15
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    188
    +0.24
    0.04
    350
    +0.18
    0.05
    219
    +0.15
    0.04
    Negative Logits
    ween
    -1.77
    omed
    -1.52
    ately
    -1.49
    endez
    -1.46
    olu
    -1.44
    otion
    -1.43
    actic
    -1.41
    iday
    -1.40
     Japan
    -1.39
    "];
    -1.39
    POSITIVE LOGITS
    İ
    3.78
    Ľ
    3.71
    ĺ
    3.54
    ŀ
    3.52
    ĸ
    3.44
    ķ
    3.42
    Ĭ
    3.40
    Ĵ
    3.40
    »
    3.38
    IJ
    3.35
    Activations Density 0.137%

    No Known Activations