© 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. 1-RES-SM
    11. 28960
    Prev
    Next
    INDEX
    Explanations

    phrases indicating locations or contexts within texts

    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
    23
    +0.16
    0.9%
    400
    +0.16
    0.9%
    300
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.16
    0.04
    300
    +0.16
    0.04
    55
    +0.13
    0.03
    Negative Logits
    documentclass
    -1.92
    visor
    -1.67
    zier
    -1.62
    watson
    -1.61
    swer
    -1.54
    mina
    -1.45
    ipore
    -1.43
    uchi
    -1.41
    arest
    -1.39
     tiempo
    -1.39
    POSITIVE LOGITS
    ķ
    2.90
    ¾
    2.86
    ¦
    2.72
    ł
    2.63
    ĥ½
    2.62
    Ĥ
    2.61
    ĺ
    2.60
    ¿
    2.56
    ģ
    2.54
    Ŀ
    2.53
    Activations Density 0.332%

    No Known Activations