Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 34350
    Prev
    Next
    INDEX
    Explanations

    historical and technical details or explanations

    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
    447
    +0.23
    1.1%
    393
    +0.14
    0.7%
    88
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.23
    0.28
    362
    +0.14
    0.16
    640
    +0.10
    0.26
    Negative Logits
    estyles
    -0.65
    istries
    -0.61
    incial
    -0.60
    icides
    -0.59
     Tracks
    -0.57
     aligned
    -0.56
    icide
    -0.56
    sworth
    -0.55
     secondly
    -0.55
    psey
    -0.55
    POSITIVE LOGITS
    animate
    0.85
    clus
    0.81
    clusively
    0.73
    liest
    0.71
    humane
    0.70
    aug
    0.70
    shore
    0.70
    organic
    0.69
    clusions
    0.68
    effect
    0.68
    Activations Density 11.058%

    No Known Activations