© 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. 9624
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the absence or negation of something

    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
    +0.21
    1.2%
    306
    +0.19
    1.1%
    348
    +0.18
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    315
    +0.21
    0.05
    52
    +0.19
    0.04
    393
    +0.18
    0.02
    Negative Logits
    iolet
    -1.58
    ighed
    -1.55
    érie
    -1.51
    ató
    -1.50
    Č
    -1.47
    ocarcin
    -1.46
    atinum
    -1.44
    aughters
    -1.43
    azine
    -1.42
    ides
    -1.39
    POSITIVE LOGITS
     regard
    2.79
     compromising
    2.24
     hesitation
    2.23
     any
    2.14
     sacrific
    2.14
     slightest
    2.06
     interruption
    1.99
     worrying
    1.99
     fuss
    1.92
     anymore
    1.87
    Activations Density 0.249%

    No Known Activations