© 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. 9818
    Prev
    Next
    INDEX
    Explanations

    elements of personal relationships and their dynamics

    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
    21
    +0.36
    2.4%
    447
    +0.13
    0.8%
    43
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    186
    +0.36
    0.12
    401
    +0.13
    0.05
    474
    +0.11
    0.14
    Negative Logits
    ↵
    -9.46
    <|outofrange|>
    -9.46
    ↵
    -9.46
    ↵↵
    -9.46
    -9.46
         ↵
    -9.46
    -9.46
     ↵
    -9.46
    ↵            ↵
    -9.46
    -9.46
    POSITIVE LOGITS
     whom
    1.68
     itself
    1.51
     whose
    1.46
    whose
    1.44
    >;
    1.42
     who
    1.41
     Himself
    1.37
     himself
    1.30
     hadn
    1.29
     herself
    1.27
    Activations Density 5.793%

    No Known Activations