© 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. 2-RES-SM
    11. 31473
    Prev
    Next
    INDEX
    Explanations

    mentions of friendship and relationships

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    406
    +0.14
    0.8%
    365
    +0.14
    0.8%
    274
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    365
    +0.14
    0.03
    406
    +0.14
    0.03
    56
    +0.12
    0.00
    Negative Logits
    ĵ
    -3.53
    ©
    -3.41
    ĩ
    -3.40
    ¯
    -3.26
    º
    -3.15
    ĨĴ
    -3.00
    Ĵ
    -3.00
    ↵
    -2.96
    -2.96
    č↵
    -2.96
    POSITIVE LOGITS
    ships
    2.81
    liest
    2.74
    liness
    2.64
    ship
    2.51
    lier
    2.14
    lies
    1.89
    stown
    1.71
    lei
    1.70
    hesis
    1.69
    doms
    1.66
    Activations Density 0.252%

    No Known Activations