© 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. 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. 826
    Prev
    Next
    INDEX
    Explanations

    phrases indicating suspicion or involvement in criminal activities

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating belief or suspicion about someone's involvement or association with an event

    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
     anyways
    -0.71
    DragonMagazine
    -0.70
     anyway
    -0.69
    Interstitial
    -0.63
    ONE
    -0.63
    ahead
    -0.62
    Products
    -0.60
     arrangements
    -0.60
     learnt
    -0.59
    !!!!!
    -0.57
    POSITIVE LOGITS
    aband
    0.68
    tips
    0.64
    ufact
    0.63
    ghai
    0.62
    isse
    0.62
    pop
    0.60
    ghazi
    0.59
    western
    0.56
     Tre
    0.56
    heading
    0.56
    Activations Density 0.286%

    No Known Activations