© 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. 32086
    Prev
    Next
    INDEX
    Explanations

    references to roles and responsibilities in various contexts, particularly related to leadership, decision-making, and the impact of actions on others

    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
    326
    +0.12
    0.6%
    475
    +0.11
    0.6%
    448
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    306
    +0.12
    0.09
    448
    +0.11
    0.04
    326
    +0.11
    0.05
    Negative Logits
    ા
    -1.73
    aler
    -1.69
    ubern
    -1.63
    à«ĩ
    -1.63
    ales
    -1.62
    anches
    -1.60
    ankind
    -1.56
    ams
    -1.53
    ะ
    -1.52
    à°¿
    -1.51
    POSITIVE LOGITS
    shot
    1.90
     been
    1.62
     pockets
    1.60
     expectation
    1.59
     clearance
    1.58
     reservations
    1.55
    ised
    1.53
     goals
    1.52
    nier
    1.52
     sung
    1.51
    Activations Density 1.728%

    No Known Activations