© 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. 14690
    Prev
    Next
    INDEX
    Explanations

    references to philosophical concepts and discussions within academic contexts

    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
    +0.18
    1.1%
    418
    +0.15
    0.9%
    400
    +0.14
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    156
    +0.18
    0.20
    111
    +0.15
    0.17
    71
    +0.14
    0.12
    Negative Logits
    »¿
    -5.11
    ĻĤ
    -5.09
    Į
    -5.00
    ª
    -4.94
    Ī
    -4.92
    ĭ
    -4.79
    ī
    -4.72
    ľĵ
    -4.71
    ¶
    -4.68
    Ĺ
    -4.66
    POSITIVE LOGITS
     conceptual
    3.12
     epistem
    2.93
     contextual
    2.71
     methodological
    2.58
     phenomenological
    2.44
     narratives
    2.39
     normative
    2.37
     critique
    2.37
     ontology
    2.37
     soci
    2.35
    Activations Density 2.972%

    No Known Activations