© 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. 31465
    Prev
    Next
    INDEX
    Explanations

    references to young people or youth-related themes

    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
    156
    +0.17
    0.9%
    423
    +0.14
    0.8%
    362
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    457
    +0.17
    0.03
    362
    +0.14
    0.03
    1
    +0.14
    0.02
    Negative Logits
     None
    -1.60
    omitempty
    -1.56
    tte
    -1.53
    acon
    -1.53
     ""
    -1.52
    oof
    -1.52
    yses
    -1.51
     '')
    -1.43
     "")
    -1.42
    ount
    -1.41
    POSITIVE LOGITS
    sters
    2.52
     adulthood
    2.17
    ster
    1.97
    stown
    1.97
    lings
    1.94
    agers
    1.81
    ł
    1.74
    hood
    1.69
    stance
    1.67
    flower
    1.66
    Activations Density 0.108%

    No Known Activations