© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 12436
    Prev
    Next
    INDEX
    Explanations

    names of key individuals or entities related to specific cultural or artistic subjects

    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
    156
    +0.29
    1.7%
    364
    +0.12
    0.7%
    77
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    383
    +0.29
    0.07
    364
    +0.12
    0.07
    307
    +0.11
    0.05
    Negative Logits
     .=
    -1.63
    eeper
    -1.57
    inates
    -1.47
    FROM
    -1.44
     behalf
    -1.43
    cedes
    -1.43
     ÑĩÑĤо
    -1.39
     hers
    -1.39
     else
    -1.39
    */
    -1.38
    POSITIVE LOGITS
    inical
    1.51
    riz
    1.48
    ¼
    1.47
    ij
    1.46
    iella
    1.44
    rael
    1.41
    ĻĤ
    1.37
    ¹
    1.36
    icia
    1.30
    uminous
    1.28
    Activations Density 1.306%

    No Known Activations