© 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. 30660
    Prev
    Next
    INDEX
    Explanations

    references to pets and their relationships with owners

    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
    407
    +0.14
    0.8%
    247
    +0.12
    0.7%
    256
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    407
    +0.14
    0.14
    403
    +0.12
    0.11
    365
    +0.12
    0.12
    Negative Logits
    ±
    -4.04
    Ŀ
    -3.94
    ¿½
    -3.91
    ķ
    -3.91
    ł
    -3.86
    ¦
    -3.85
    ·
    -3.83
    ¿
    -3.82
    ĻĤ
    -3.81
    ª
    -3.77
    POSITIVE LOGITS
    ressor
    1.79
    ourse
    1.75
     deserves
    1.74
    breaker
    1.59
    aña
    1.54
    lookup
    1.49
    scape
    1.46
    checks
    1.46
     includes
    1.45
    Compat
    1.45
    Activations Density 3.581%

    No Known Activations