© 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. 13422
    Prev
    Next
    INDEX
    Explanations

    terms related to software usage rights and permissions

    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
    320
    +0.16
    0.9%
    47
    +0.14
    0.8%
    323
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    222
    +0.16
    0.02
    320
    +0.14
    0.03
    74
    +0.11
    0.02
    Negative Logits
    ī
    -3.29
    Ĺ
    -3.25
    §
    -3.25
    Į
    -3.22
    Ī
    -3.19
    ı
    -3.17
    į
    -3.09
    ®
    -3.09
    <|endoftext|>
    -3.05
    ¯
    -3.05
    POSITIVE LOGITS
     freely
    1.58
     molecules
    1.53
     therefrom
    1.47
    imiento
    1.46
     thereon
    1.40
     activities
    1.39
     portions
    1.38
     decom
    1.36
    ricular
    1.36
    aceous
    1.36
    Activations Density 0.146%

    No Known Activations