© 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. 14777
    Prev
    Next
    INDEX
    Explanations

    phrases indicating relationships or associations between variables

    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
    307
    +0.19
    1.1%
    306
    +0.17
    1.0%
    63
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    306
    +0.19
    0.02
    307
    +0.17
    0.03
    63
    +0.13
    0.01
    Negative Logits
    worth
    -1.71
    mans
    -1.60
    vier
    -1.60
    stown
    -1.56
     exploits
    -1.53
    man
    -1.48
    RICT
    -1.43
    abbit
    -1.42
    \\
    -1.39
    crets
    -1.38
    POSITIVE LOGITS
    ¶
    4.69
    ·
    4.69
    »¿
    4.63
    ¹
    4.61
    4.42
    4.42
    ↵
    4.42
    4.42
    4.42
    4.42
    Activations Density 0.137%

    No Known Activations