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

    Neuronpedia

    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. 3-RES-JB
    11. 1797
    Prev
    Next
    INDEX
    Explanations

    phrases related to advising or urging against certain actions

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to abstaining or refraining from actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ammy
    -1.01
    NetMessage
    -0.75
    immer
    -0.70
    rations
    -0.69
    ramid
    -0.69
    ovies
    -0.67
    neau
    -0.67
    odes
    -0.67
    oÄŁ
    -0.66
    onomy
    -0.65
    POSITIVE LOGITS
     refrain
    1.18
    rences
    0.90
     abst
    0.86
    SourceFile
    0.78
    ////////
    0.69
    ministic
    0.69
     stren
    0.67
     answering
    0.67
     acknow
    0.66
    swer
    0.66
    Activations Density 0.008%

    No Known Activations