© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 3765
    Prev
    Next
    INDEX
    Explanations

    words related to government, security, and political disputes

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    216
    +0.10
    0.6%
    276
    +0.08
    0.5%
    466
    +0.08
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.21
    216
    +0.08
    0.10
    48
    +0.08
    0.10
    Negative Logits
    swick
    -0.55
    ãĥ©ãĥ³
    -0.51
    INESS
    -0.50
    desktop
    -0.49
    usage
    -0.48
    nesday
    -0.47
     McDonnell
    -0.47
    clair
    -0.47
     visitor
    -0.46
    case
    -0.46
    POSITIVE LOGITS
    aging
    0.60
    oti
    0.59
    agement
    0.58
    pton
    0.57
    adra
    0.56
    agos
    0.56
    ests
    0.52
     Cyborg
    0.52
    kees
    0.52
    age
    0.52
    Activations Density 7.328%

    No Known Activations