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. 11935
    Prev
    Next
    INDEX
    Explanations

    the word "who" and its variations, indicating a focus on questioning identities or relationships

    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
    206
    +0.15
    0.8%
    23
    +0.14
    0.8%
    204
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    206
    +0.15
    0.06
    356
    +0.14
    0.05
    434
    +0.12
    0.05
    Negative Logits
    Category
    -1.77
    ees
    -1.73
    ITED
    -1.65
    ITION
    -1.62
     itself
    -1.61
    ľ
    -1.59
    CRIPTION
    -1.59
    ŀ
    -1.51
    eners
    -1.50
    ļ
    -1.49
    POSITIVE LOGITS
    oping
    2.07
    pping
    1.99
    sell
    1.68
    osh
    1.68
    oped
    1.63
    pper
    1.59
    opsies
    1.59
    thon
    1.59
    haul
    1.58
     else
    1.57
    Activations Density 0.114%

    No Known Activations