© 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. 32072
    Prev
    Next
    INDEX
    Explanations

    mentions of file paths or references to file dependencies

    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
    386
    +0.12
    0.6%
    23
    +0.11
    0.6%
    38
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    425
    +0.12
    0.05
    53
    +0.11
    0.04
    318
    +0.11
    0.02
    Negative Logits
    ij
    -3.37
    ĻĤ
    -3.13
    Ģ
    -3.06
    ĨĴ
    -2.98
    ĥ
    -2.87
    §
    -2.85
    ¡
    -2.82
    ľĵ
    -2.81
    Ĵ
    -2.78
    Ń
    -2.73
    POSITIVE LOGITS
    fected
    1.40
     flesh
    1.37
    jd
    1.30
    ften
    1.29
    }{-
    1.29
     bare
    1.25
    ember
    1.24
     delegation
    1.24
    fections
    1.23
    TRODUCTION
    1.22
    Activations Density 0.078%

    No Known Activations