© 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. 1-RES-SM
    11. 29846
    Prev
    Next
    INDEX
    Explanations

    phrases about inclusivity and universality

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    261
    +0.13
    0.7%
    159
    +0.12
    0.7%
    505
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    261
    +0.13
    0.07
    362
    +0.12
    0.06
    372
    +0.12
    0.05
    Negative Logits
    oddsidemargin
    -1.91
    scale
    -1.51
     quo
    -1.50
    backed
    -1.50
    stood
    -1.49
     toward
    -1.48
     limits
    -1.45
     normalized
    -1.45
    opacity
    -1.44
    posted
    -1.43
    POSITIVE LOGITS
    ĸ´
    2.62
    »¿
    2.52
    ¼
    2.50
    ľ
    2.46
    ¿½
    2.37
    Ł
    2.31
    ĭ
    2.30
    ·¸
    2.29
    ķ
    2.27
    ¤
    2.26
    Activations Density 0.482%

    No Known Activations