© 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. 30931
    Prev
    Next
    INDEX
    Explanations

    phrases related to statistical analysis or results

    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
    136
    +0.19
    1.0%
    412
    +0.13
    0.7%
    120
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    136
    +0.19
    0.00
    208
    +0.13
    0.04
    128
    +0.11
    0.02
    Negative Logits
    ±
    -3.65
    Ļ
    -3.55
    ħ
    -3.16
    Ļª
    -3.11
    İ
    -3.10
    «
    -3.09
    ´
    -3.07
    ²
    -3.06
    ĺ
    -3.06
    Ħ
    -3.03
    POSITIVE LOGITS
    ){#
    1.79
    ipolar
    1.69
    cel
    1.44
    uffled
    1.43
    pled
    1.41
    .).
    1.41
     harmed
    1.40
    suc
    1.40
    imester
    1.40
     unfavorable
    1.35
    Activations Density 0.124%

    No Known Activations