Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 29554
    Prev
    Next
    INDEX
    Explanations

    expressions of acceptable social behavior or moral standing

    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
    198
    +0.24
    1.5%
    400
    +0.23
    1.4%
    186
    +0.20
    1.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    400
    +0.24
    0.10
    198
    +0.23
    0.13
    157
    +0.20
    0.14
    Negative Logits
    onica
    -1.64
    quart
    -1.63
     Archives
    -1.56
    pora
    -1.52
     Awards
    -1.51
    éric
    -1.50
    chaft
    -1.42
     Edited
    -1.40
     offices
    -1.37
    amycin
    -1.36
    POSITIVE LOGITS
     won
    1.60
    addy
    1.50
    ections
    1.49
    ickets
    1.48
    &&
    1.48
     ![
    1.43
     anything
    1.38
    osition
    1.38
    olver
    1.37
    exists
    1.36
    Activations Density 3.161%

    No Known Activations