© 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 EvalsExportsAPI 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. 10837
    Prev
    Next
    INDEX
    Explanations

    frequent personal pronouns and expressions of self-reference

    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
    209
    +0.19
    1.1%
    301
    +0.14
    0.8%
    71
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    301
    +0.19
    0.08
    171
    +0.14
    0.06
    498
    +0.12
    0.06
    Negative Logits
    »¿
    -3.55
    Ħ
    -3.36
    ¼
    -3.27
    ij
    -3.23
    ŀ
    -3.17
    Īĺ
    -3.11
    ª
    -2.99
    Ļ
    -2.98
    º
    -2.96
    ł
    -2.92
    POSITIVE LOGITS
     am
    3.10
     myself
    2.62
     believe
    2.28
     have
    2.27
     my
    2.23
     intend
    2.13
     want
    2.11
     understand
    2.08
     know
    1.98
     hope
    1.96
    Activations Density 0.220%

    No Known Activations