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

    Neuronpedia

    APICircuit 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. 4455
    Prev
    Next
    INDEX
    Explanations

    opportunities or advantageous situations in a text, such as exploiting certain conditions for personal gain or benefit

    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
    +0.21
    1.2%
    679
    +0.12
    0.7%
    288
    +0.08
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.21
    0.25
    338
    +0.12
    0.10
    743
    +0.08
    0.05
    Negative Logits
    sis
    -0.57
    don
    -0.56
    gon
    -0.51
    jah
    -0.51
    liam
    -0.51
     brace
    -0.50
    dot
    -0.50
    odd
    -0.49
    mad
    -0.46
    rib
    -0.46
    POSITIVE LOGITS
    fully
    0.63
    ously
    0.60
    ful
    0.58
    uristic
    0.57
    isance
    0.55
    oise
    0.54
    ibility
    0.54
    enment
    0.54
    udo
    0.53
    ileged
    0.53
    Activations Density 12.443%

    No Known Activations