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. 13468
    Prev
    Next
    INDEX
    Explanations

    descriptions of violent encounters and criminal activities, particularly involving physical harm

    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
    +0.27
    1.6%
    393
    +0.17
    1.0%
    36
    +0.16
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.27
    0.39
    393
    +0.17
    0.29
    159
    +0.16
    0.18
    Negative Logits
     relies
    -0.62
    Rules
    -0.61
    Edit
    -0.60
    Own
    -0.60
    Events
    -0.60
    evidence
    -0.59
    Based
    -0.58
    Sources
    -0.58
    alties
    -0.58
     amounts
    -0.58
    POSITIVE LOGITS
     handful
    0.88
     bunch
    0.86
     flurry
    0.82
     bottle
    0.80
     couple
    0.80
     knife
    0.76
    rouse
    0.76
     small
    0.75
     few
    0.75
     dozen
    0.74
    Activations Density 10.275%

    No Known Activations