© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 33128
    Prev
    Next
    INDEX
    Explanations

    words related to historical figures, locations, and events

    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
    71
    +0.06
    0.6%
    587
    +0.04
    0.4%
    725
    +0.04
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.06
    0.11
    111
    +0.04
    0.08
    71
    +0.04
    -0.02
    Negative Logits
    izable
    -0.43
    zees
    -0.37
    Bird
    -0.31
    uscript
    -0.30
    cel
    -0.30
    apult
    -0.30
    quest
    -0.30
    izes
    -0.30
    Pear
    -0.30
    Lot
    -0.30
    POSITIVE LOGITS
    ARDS
    0.54
    ARD
    0.54
    furt
    0.43
    NESS
    0.43
    ãģį
    0.42
    bourg
    0.41
    ERE
    0.41
    sburgh
    0.41
    sburg
    0.39
    ERAL
    0.38
    Activations Density 19.203%

    No Known Activations