© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 1-RES-SM
    11. 27585
    Prev
    Next
    INDEX
    Explanations

    instances of the word "Further"

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    464
    +0.13
    0.7%
    23
    +0.11
    0.6%
    84
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    464
    +0.13
    0.02
    84
    +0.11
    0.01
    452
    +0.10
    0.02
    Negative Logits
    ĩ
    -2.01
    ĥ½
    -1.74
    aches
    -1.72
    Ģ
    -1.63
    ellers
    -1.63
    Ĥ¬
    -1.60
    eller
    -1.54
    kered
    -1.53
    ½
    -1.52
    º
    -1.51
    POSITIVE LOGITS
    going
    1.73
    where
    1.66
     Article
    1.56
     junction
    1.43
     explanation
    1.42
     Justice
    1.36
    what
    1.35
     appreciation
    1.34
    ming
    1.34
     freedom
    1.34
    Activations Density 0.050%

    No Known Activations