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. 11840
    Prev
    Next
    INDEX
    Explanations

    elements of legal language and compliance statements

    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
    382
    +0.12
    0.7%
    323
    +0.12
    0.7%
    222
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    192
    +0.12
    0.03
    222
    +0.12
    0.03
    242
    +0.10
    0.03
    Negative Logits
    mouth
    -1.47
    thon
    -1.40
    PCs
    -1.39
    ľĵ
    -1.34
    erialize
    -1.31
     worldwide
    -1.27
     fans
    -1.25
    $)
    -1.24
    oret
    -1.24
     lovers
    -1.24
    POSITIVE LOGITS
    sible
    1.57
     contempt
    1.38
    ане
    1.32
    ------------------------------------------------
    1.32
    iousness
    1.29
    iedad
    1.28
    ----------------------------------------------------------------
    1.28
    ius
    1.28
    iative
    1.28
     ---
    1.26
    Activations Density 0.069%

    No Known Activations