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. 29342
    Prev
    Next
    INDEX
    Explanations

    comments or documentation sections in code

    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
    240
    +0.13
    0.7%
    410
    +0.12
    0.7%
    478
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    411
    +0.13
    0.03
    165
    +0.12
    0.03
    463
    +0.12
    0.02
    Negative Logits
    ata
    -1.57
    ight
    -1.55
    ath
    -1.54
     side
    -1.52
    bow
    -1.47
    rish
    -1.45
    shire
    -1.43
    sid
    -1.36
    oi
    -1.36
    TON
    -1.35
    POSITIVE LOGITS
    ¿½
    1.92
    ½
    1.59
    headed
    1.56
     ourselves
    1.53
    inement
    1.52
     (@
    1.52
    :--
    1.50
    cases
    1.48
    ĥ½
    1.45
    formance
    1.44
    Activations Density 0.068%

    No Known Activations