© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 15013
    Prev
    Next
    INDEX
    Explanations

    keywords or phrases related to caution or health warnings

    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
    307
    +0.33
    2.0%
    306
    +0.13
    0.8%
    23
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.33
    0.06
    46
    +0.13
    0.06
    267
    +0.13
    0.05
    Negative Logits
    »¿
    -2.18
    ership
    -2.12
    ŀ
    -1.92
    etheless
    -1.60
    Į
    -1.55
    Ļª
    -1.54
    İ
    -1.50
    ·¸
    -1.49
    Ľ
    -1.48
     cui
    -1.48
    POSITIVE LOGITS
     compared
    2.42
     confronted
    2.14
     they
    2.05
     asked
    2.04
     viewed
    1.89
     dealing
    1.86
    fy
    1.78
     considering
    1.76
     faced
    1.76
     necessary
    1.75
    Activations Density 0.240%

    No Known Activations