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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 45981
    Prev
    Next
    INDEX
    Explanations

    website URLs and potentially email addresses

    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
    246
    +0.03
    0.5%
    746
    +0.03
    0.4%
    277
    +0.03
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    283
    +0.03
    0.02
    640
    +0.03
    0.17
    393
    +0.03
    0.18
    Negative Logits
    grave
    -0.24
     Exhibit
    -0.21
    Forward
    -0.18
     Revival
    -0.17
     Cullen
    -0.17
     Grave
    -0.17
     Travels
    -0.17
    template
    -0.17
     Gemini
    -0.17
    OME
    -0.17
    POSITIVE LOGITS
    amac
    0.27
    ording
    0.21
    aser
    0.21
    oshop
    0.21
    abin
    0.20
    apon
    0.20
    eret
    0.20
    izophren
    0.20
    orno
    0.19
     whim
    0.18
    Activations Density 5.452%

    No Known Activations