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. 11487
    Prev
    Next
    INDEX
    Explanations

    words related to tiles and flooring materials

    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
    368
    +0.19
    1.1%
    232
    +0.13
    0.7%
    434
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.19
    0.03
    368
    +0.13
    0.01
    232
    +0.13
    0.03
    Negative Logits
    illi
    -1.50
    fections
    -1.49
    ique
    -1.42
    fected
    -1.41
    orems
    -1.41
     living
    -1.34
     office
    -1.32
    imester
    -1.30
    chro
    -1.27
     teaching
    -1.26
    POSITIVE LOGITS
    į
    3.38
    Ģ
    3.32
    ĩ
    3.00
    ¯
    2.93
    ¢
    2.86
    ģ
    2.84
    ¿
    2.83
    ŀ
    2.79
    ĥ
    2.67
    ¶
    2.66
    Activations Density 0.235%

    No Known Activations