© 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. 8913
    Prev
    Next
    INDEX
    Explanations

    references to specific groups or roles within various contexts, such as professions and identities

    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
    450
    +0.14
    0.8%
    380
    +0.13
    0.7%
    306
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    380
    +0.14
    0.13
    242
    +0.13
    0.11
    229
    +0.11
    0.12
    Negative Logits
    ĻĤ
    -4.60
    į
    -4.37
     ↵
    -4.34
    č↵
    -4.33
    ↵
    -4.33
    ↵  ³³³
    -4.33
    ↵
    -4.33
              ↵
    -4.33
    -4.33
    <|padding|>
    -4.33
    POSITIVE LOGITS
    hip
    2.55
     prefer
    2.25
     have
    2.24
     want
    2.17
    pace
    2.12
     contend
    2.11
     themselves
    2.11
     argue
    2.08
     tend
    2.05
     flock
    2.02
    Activations Density 1.140%

    No Known Activations