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. 34339
    Prev
    Next
    INDEX
    Explanations

    terms coined for new concepts or ideas, potentially related to innovative or novel language usage

    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
    578
    +0.09
    0.6%
    187
    +0.06
    0.5%
    391
    +0.06
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.09
    0.17
    608
    +0.06
    0.22
    526
    +0.06
    -0.10
    Negative Logits
    endi
    -0.42
    iment
    -0.38
    estate
    -0.38
    room
    -0.37
    ittal
    -0.36
    finder
    -0.35
    emb
    -0.35
     Clintons
    -0.35
     quo
    -0.34
     strongh
    -0.34
    POSITIVE LOGITS
    ãĥ¼ãĥ«
    0.52
    pmwiki
    0.48
     coined
    0.47
    ãĥĥãĥĪ
    0.47
    rences
    0.47
    unci
    0.47
    forth
    0.46
    orne
    0.46
    itialized
    0.45
    vision
    0.45
    Activations Density 14.828%

    No Known Activations