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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 11046
    Prev
    Next
    INDEX
    Explanations

    phrases expressing opinions or beliefs

    oai_token-act-pair · gpt-3.5-turbo

    expressions of disagreement or negative opinions on controversial topics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ubb
    -0.80
    ffen
    -0.68
    ran
    -0.64
     manoeuv
    -0.63
    çīĪ
    -0.63
    iculty
    -0.62
    rar
    -0.61
    icro
    -0.60
    OME
    -0.60
    rero
    -0.60
    POSITIVE LOGITS
     hence
    0.89
     whereas
    0.88
     therefore
    0.85
     implying
    0.83
    Therefore
    0.83
    because
    0.79
     thus
    0.79
     secondly
    0.77
     blah
    0.77
    rather
    0.76
    Activations Density 0.613%

    No Known Activations