© 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 EvalsExportsAPI 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. 12964
    Prev
    Next
    INDEX
    Explanations

    phrases related to venues and event spaces

    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
    25
    +0.21
    1.2%
    156
    +0.19
    1.1%
    466
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    25
    +0.21
    0.20
    58
    +0.19
    0.18
    466
    +0.13
    0.15
    Negative Logits
    ī
    -6.07
    Ĥ¬
    -5.63
    Ī
    -5.38
    Į
    -5.16
    ĥ½
    -5.13
    ²
    -5.05
    §
    -4.97
    ¤
    -4.97
    Ĥ
    -4.96
    ľĵ
    -4.96
    POSITIVE LOGITS
     etc
    1.82
     ecology
    1.68
     social
    1.66
     respectively
    1.66
    ensory
    1.64
     Anthrop
    1.64
     anthrop
    1.63
     Humans
    1.61
     Objects
    1.60
     Consc
    1.59
    Activations Density 4.524%

    No Known Activations