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. 11499
    Prev
    Next
    INDEX
    Explanations

    instances of the word "currently."

    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
    246
    +0.14
    0.8%
    365
    +0.11
    0.6%
    193
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    155
    +0.14
    0.02
    246
    +0.11
    0.02
    483
    +0.10
    0.02
    Negative Logits
    ĻĤ
    -3.94
    Ļª
    -3.81
    ĨĴ
    -3.40
    ª
    -3.28
    Ĵ
    -3.23
    Ħ
    -3.16
    ģ
    -3.13
    ĭ
    -3.05
    ©
    -2.96
    ĩ
    -2.95
    POSITIVE LOGITS
     underway
    2.21
    here
    1.81
    far
    1.80
    adays
    1.75
     hunted
    1.68
    occupied
    1.61
     occupied
    1.61
    held
    1.60
     marketed
    1.56
     undergoing
    1.55
    Activations Density 0.043%

    No Known Activations