© 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. 14622
    Prev
    Next
    INDEX
    Explanations

    references to air-related concepts and conditions

    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
    164
    +0.13
    0.7%
    443
    +0.12
    0.7%
    395
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    164
    +0.13
    0.04
    275
    +0.12
    0.02
    246
    +0.12
    0.03
    Negative Logits
    nis
    -1.46
    aurus
    -1.44
     handwriting
    -1.44
     mind
    -1.43
     honour
    -1.42
     autosomal
    -1.42
    teen
    -1.41
    imab
    -1.40
    ate
    -1.39
    omerase
    -1.37
    POSITIVE LOGITS
    planes
    2.69
    flow
    2.65
    Ĵ
    2.64
    tight
    2.38
    waves
    2.25
    ŀ
    2.23
    »
    2.20
     conditioning
    2.19
    č↵
    2.18
    2.18
    Activations Density 0.110%

    No Known Activations