© 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. 13632
    Prev
    Next
    INDEX
    Explanations

    mentions of legal incidents, criminal activities, conflicts with authority, and social or political debates

    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
    167
    +0.13
    0.5%
    447
    +0.10
    0.4%
    647
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    154
    +0.13
    0.10
    308
    +0.10
    0.07
    199
    +0.10
    0.11
    Negative Logits
     destro
    -0.79
    wagen
    -0.75
     potatoes
    -0.74
     guiActiveUnfocused
    -0.74
    oulos
    -0.73
    terday
    -0.72
     finance
    -0.72
    ORED
    -0.72
     Clarkson
    -0.70
     potato
    -0.70
    POSITIVE LOGITS
    xt
    1.10
    iced
    1.08
    iph
    1.06
    idd
    1.05
    ------------
    1.05
    lli
    1.05
    Ëľ
    1.05
    urn
    1.05
    ickle
    1.04
    llo
    1.03
    Activations Density 3.330%

    No Known Activations