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. 29424
    Prev
    Next
    INDEX
    Explanations

    conjunctions and phrases that express connection or addition

    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
    478
    +0.19
    1.1%
    23
    +0.12
    0.7%
    454
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    402
    +0.19
    0.11
    146
    +0.12
    0.11
    494
    +0.11
    0.09
    Negative Logits
    stown
    -1.88
     yourself
    -1.81
    feit
    -1.72
    mate
    -1.54
    ungs
    -1.53
     yourselves
    -1.53
    matter
    -1.53
     himself
    -1.52
    idium
    -1.50
    acher
    -1.49
    POSITIVE LOGITS
     hence
    1.92
     consequent
    1.85
    /)
    1.74
     subsequent
    1.72
    WHM
    1.63
     consequently
    1.59
     eventual
    1.59
     yes
    1.56
     occasional
    1.55
    edes
    1.50
    Activations Density 1.062%

    No Known Activations