© 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. 4-RES-SM
    11. 14613
    Prev
    Next
    INDEX
    Explanations

    articles and quantifiers related to individuals or objects

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    +0.19
    1.1%
    137
    +0.17
    1.0%
    145
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    137
    +0.19
    0.14
    145
    +0.17
    0.10
    291
    +0.14
    0.08
    Negative Logits
    ophores
    -1.82
    iest
    -1.76
    ERTY
    -1.69
     whatsoever
    -1.67
    ário
    -1.60
    exports
    -1.55
    umber
    -1.54
    rex
    -1.54
    rums
    -1.51
     except
    -1.51
    POSITIVE LOGITS
     dozen
    2.53
     different
    2.23
     massive
    2.23
     recent
    2.22
     huge
    2.19
     series
    2.16
     similar
    2.14
     nearby
    2.07
     horrific
    2.01
     handful
    1.99
    Activations Density 0.840%

    No Known Activations