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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 29049
    Prev
    Next
    INDEX
    Explanations

    conjunctions and alternative phrases indicating choice or options

    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
    436
    +0.14
    0.8%
    350
    +0.13
    0.7%
    261
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    50
    +0.14
    0.05
    494
    +0.13
    0.04
    347
    +0.12
    0.04
    Negative Logits
    ump
    -1.56
    watson
    -1.55
    xspace
    -1.49
    opts
    -1.47
    MD
    -1.32
    aryng
    -1.32
    unge
    -1.31
    arynge
    -1.29
    verted
    -1.29
    ursuant
    -1.28
    POSITIVE LOGITS
    naments
    1.71
    chard
    1.70
     contemplate
    1.42
     rue
    1.39
     equivalently
    1.35
    ¿
    1.34
    vier
    1.30
     least
    1.27
     Trustee
    1.27
     collaborators
    1.26
    Activations Density 0.240%

    No Known Activations