© 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. 10611
    Prev
    Next
    INDEX
    Explanations

    phrases that specify comparisons or characteristics

    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
    +0.31
    1.9%
    348
    +0.13
    0.8%
    478
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    144
    +0.31
    0.06
    2
    +0.13
    0.05
    361
    +0.12
    0.06
    Negative Logits
    ľĵ
    -2.10
    »¿
    -2.05
    ±
    -2.00
    Ĩ
    -1.95
    §
    -1.93
    Ģ
    -1.88
    Ħ
    -1.84
    ħ
    -1.83
    «
    -1.83
    į
    -1.83
    POSITIVE LOGITS
     follows
    2.28
    phalt
    1.92
    pired
    1.82
    crib
    1.78
    inine
    1.75
     having
    1.74
     being
    1.74
     part
    1.72
     belonging
    1.72
    sembles
    1.72
    Activations Density 0.274%

    No Known Activations