© 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. 9740
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate preparatory actions or situations prior to events

    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
    +0.28
    1.7%
    466
    +0.15
    0.9%
    306
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    315
    +0.28
    0.05
    91
    +0.15
    0.05
    101
    +0.14
    0.04
    Negative Logits
    į
    -2.60
    Ģ
    -2.35
    ±
    -2.24
    ħ
    -2.24
    Ħ
    -2.03
    ·¸
    -1.96
    Į
    -1.93
    Ĵ
    -1.89
    ĩ
    -1.81
    ¡
    -1.81
    POSITIVE LOGITS
     proceeding
    2.33
     embark
    2.13
     entering
    2.07
     exiting
    1.99
     submitting
    1.97
     settling
    1.92
     dawn
    1.91
     deciding
    1.90
     retiring
    1.88
     transferring
    1.86
    Activations Density 0.237%

    No Known Activations