© 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. 27256
    Prev
    Next
    INDEX
    Explanations

    expressions of personal or subjective experience

    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
    469
    +0.14
    0.8%
    278
    +0.12
    0.7%
    227
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    363
    +0.14
    0.07
    184
    +0.12
    0.05
    469
    +0.12
    0.06
    Negative Logits
    casting
    -1.74
    mium
    -1.63
    casters
    -1.58
    iply
    -1.53
    uate
    -1.52
    limiting
    -1.51
     notation
    -1.42
    ribute
    -1.41
    getInstance
    -1.41
     vary
    -1.40
    POSITIVE LOGITS
    ½
    2.73
    §
    2.50
    Īĺ
    2.35
    athered
    2.35
    ¸
    2.26
    «
    2.24
    ģ
    2.23
    Ļ
    2.21
    Ľ
    2.20
    ©
    2.19
    Activations Density 0.989%

    No Known Activations