© 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. 9808
    Prev
    Next
    INDEX
    Explanations

    directives related to data management and operations

    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
    78
    +0.21
    1.2%
    237
    +0.12
    0.7%
    240
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    334
    +0.21
    0.01
    78
    +0.12
    0.00
    255
    +0.10
    -0.00
    Negative Logits
    ativity
    -1.67
    ity
    -1.41
    ateness
    -1.41
    ating
    -1.37
     remaining
    -1.37
    ately
    -1.36
    entric
    -1.33
     loss
    -1.32
    atively
    -1.29
     deficit
    -1.29
    POSITIVE LOGITS
    İ
    3.07
    ģ
    2.86
    ¸
    2.79
    ª
    2.77
    ¤
    2.76
    º
    2.71
    ¿
    2.70
    Ĵ
    2.69
    ŀ
    2.67
    ½
    2.66
    Activations Density 0.055%

    No Known Activations