© 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. 13960
    Prev
    Next
    INDEX
    Explanations

    occurrences of the name "Don."

    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
    284
    +0.14
    0.8%
    406
    +0.11
    0.6%
    1
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    406
    +0.14
    0.04
    1
    +0.11
    0.03
    409
    +0.10
    0.04
    Negative Logits
    LES
    -1.66
    LET
    -1.61
    £
    -1.59
     ranging
    -1.57
    eon
    -1.52
     validity
    -1.51
    oler
    -1.47
     milliseconds
    -1.46
    Ĺ
    -1.45
    Ķ
    -1.44
    POSITIVE LOGITS
    't
    2.51
    ovan
    2.35
    caster
    2.08
    nie
    2.04
    avir
    2.02
    uts
    1.90
    ogh
    1.82
    dern
    1.82
    nar
    1.82
    née
    1.82
    Activations Density 0.115%

    No Known Activations