© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 9621
    Prev
    Next
    INDEX
    Explanations

    instances of the word "inter" in various forms and contexts

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    487
    +0.13
    0.8%
    98
    +0.13
    0.7%
    112
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    487
    +0.13
    0.02
    382
    +0.13
    0.01
    393
    +0.10
    0.01
    Negative Logits
    ILY
    -1.66
    ership
    -1.51
     Returns
    -1.48
    gery
    -1.46
    ful
    -1.44
    ew
    -1.43
    ering
    -1.43
    ocs
    -1.38
    ers
    -1.38
    otine
    -1.36
    POSITIVE LOGITS
     alia
    2.30
    ferences
    1.86
    facing
    1.84
    species
    1.78
    cession
    1.74
    upt
    1.73
    molecular
    1.70
    ceptor
    1.69
    cellular
    1.69
    changes
    1.68
    Activations Density 0.026%

    No Known Activations