© 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. 3670
    Prev
    Next
    INDEX
    Explanations

    mentions of specific geographical locations and historical events

    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
    373
    +0.07
    0.5%
    566
    +0.07
    0.5%
    0
    +0.07
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    577
    +0.07
    0.13
    314
    +0.07
    -0.04
    326
    +0.07
    0.22
    Negative Logits
    anism
    -0.54
    fit
    -0.48
    ions
    -0.47
    pool
    -0.45
    sets
    -0.43
    aline
    -0.42
    inet
    -0.42
    fitted
    -0.42
    atu
    -0.42
    waves
    -0.41
    POSITIVE LOGITS
    ILY
    0.59
    ICAN
    0.58
    ALLY
    0.54
    JD
    0.48
    ocard
    0.48
    ARS
    0.47
    olen
    0.46
    UFC
    0.46
    OGR
    0.46
    ORGE
    0.46
    Activations Density 7.631%

    No Known Activations