© 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. 11364
    Prev
    Next
    INDEX
    Explanations

    expressions of personal preferences or emotions regarding various topics

    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
    307
    +0.15
    0.9%
    156
    +0.12
    0.7%
    54
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    56
    +0.15
    0.03
    377
    +0.12
    0.08
    307
    +0.11
    0.11
    Negative Logits
    İ
    -2.51
    «
    -2.42
    ¿½
    -2.38
    ¸
    -2.35
    Ħ
    -2.32
    ĩ
    -2.31
    ģ
    -2.28
    Ģ
    -2.27
    ↵
    -2.27
    ↵
    -2.27
    POSITIVE LOGITS
     my
    2.82
     myself
    2.69
     me
    2.30
     I
    2.21
    _."
    2.10
    mys
    2.05
    I
    2.02
     mine
    2.02
    ")]
    2.01
    ,'"
    1.89
    Activations Density 2.096%

    No Known Activations