© 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. 4-RES-SM
    11. 13972
    Prev
    Next
    INDEX
    Explanations

    terms related to warranties and disclaimers

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    78
    +0.17
    1.1%
    420
    +0.10
    0.6%
    507
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    334
    +0.17
    0.01
    420
    +0.10
    0.01
    11
    +0.10
    0.01
    Negative Logits
    ppler
    -1.77
    ket
    -1.64
    ainen
    -1.50
    ffen
    -1.50
    hthal
    -1.46
    kin
    -1.41
    chle
    -1.40
    chaft
    -1.40
    ch
    -1.40
    aset
    -1.39
    POSITIVE LOGITS
    ŀ
    5.12
    ¤
    4.85
    4.76
    č↵
    4.76
    <|outofrange|>
    4.76
    4.76
    <|outofrange|>
    4.76
    4.76
    4.76
    č↵
    4.76
    Activations Density 0.032%

    No Known Activations