© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 26810
    Prev
    Next
    INDEX
    Explanations

    references to current events or topics of ongoing relevance

    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
    235
    +0.18
    1.1%
    240
    +0.15
    0.9%
    326
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    235
    +0.18
    0.08
    5
    +0.15
    0.07
    240
    +0.14
    -0.02
    Negative Logits
    Īĺ
    -3.35
    ĭ
    -2.99
    ĥ
    -2.83
    İ
    -2.71
    ²
    -2.59
    ľĵ
    -2.57
    ¨
    -2.55
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    -2.55
    ↵           
    -2.50
                                            
    -2.50
    POSITIVE LOGITS
    ement
    1.78
    iors
    1.49
    "_
    1.40
    iol
    1.36
    rbrace
    1.33
    ."_
    1.31
     point
    1.30
    ordinary
    1.29
    PP
    1.26
    liest
    1.24
    Activations Density 1.833%

    No Known Activations