© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 2-RES-SM
    11. 30159
    Prev
    Next
    INDEX
    Explanations

    references to alternative options or additional elements in various contexts

    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
    419
    +0.19
    1.0%
    44
    +0.13
    0.7%
    395
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    44
    +0.19
    0.06
    230
    +0.13
    0.04
    395
    +0.12
    0.06
    Negative Logits
    ]{};
    -2.14
    $]{};
    -1.83
    asure
    -1.57
    imeter
    -1.51
    ];
    -1.51
    rente
    -1.50
    itone
    -1.49
    ]{\
    -1.48
    >();
    -1.39
    rence
    -1.39
    POSITIVE LOGITS
     alike
    2.86
    ĵ
    2.05
    Ŀ
    1.97
    ħ
    1.80
     coworkers
    1.72
    ¿½
    1.68
    º
    1.68
     collaborators
    1.63
    ĸ
    1.63
     descendants
    1.63
    Activations Density 0.481%

    No Known Activations