© 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. 1-RES-SM
    11. 27741
    Prev
    Next
    INDEX
    Explanations

    legal terms and conditions related to licensing agreements

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    231
    +0.11
    0.6%
    178
    +0.10
    0.6%
    303
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    26
    +0.11
    0.03
    252
    +0.10
    0.03
    469
    +0.10
    0.04
    Negative Logits
    Ļª
    -2.32
    ĻĤ
    -2.25
    noreply
    -2.17
    ¯
    -1.96
    ¦
    -1.93
    ¼
    -1.85
    Ħ
    -1.69
    ļ
    -1.69
    ·
    -1.59
    ľĵ
    -1.56
    POSITIVE LOGITS
    fort
    1.82
    arty
    1.59
     hereby
    1.59
     illegally
    1.50
     Copyright
    1.49
     IMPLIED
    1.45
    ancial
    1.45
     entitled
    1.43
    utions
    1.40
     distributed
    1.40
    Activations Density 0.407%

    No Known Activations