© 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. 27042
    Prev
    Next
    INDEX
    Explanations

    references to legal cases and decisions

    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
    104
    +0.13
    0.7%
    237
    +0.12
    0.6%
    12
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    418
    +0.13
    -0.02
    365
    +0.12
    0.09
    237
    +0.11
    0.11
    Negative Logits
     Profile
    -1.56
    kil
    -1.50
    née
    -1.47
     nutrients
    -1.46
     Features
    -1.46
    seed
    -1.42
     migrate
    -1.42
    obi
    -1.41
    ids
    -1.38
     Generation
    -1.37
    POSITIVE LOGITS
     appellate
    2.39
     justices
    2.33
     unanimous
    2.23
     courtroom
    2.22
     precedent
    2.22
     justice
    2.21
     jur
    2.13
     judicial
    2.12
     Appeals
    2.11
     Appellate
    2.08
    Activations Density 2.915%

    No Known Activations