© 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. 1-RES-JB
    11. 847
    Prev
    Next
    INDEX
    Explanations

    phrases indicating consequences or reasons for specific actions

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate causation or reasons behind events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    abbling
    -0.70
    drive
    -0.69
     jewels
    -0.67
     sheep
    -0.67
     Sheep
    -0.66
    amped
    -0.66
    amps
    -0.65
     Herz
    -0.64
    IFT
    -0.61
    aps
    -0.60
    POSITIVE LOGITS
     simultane
    0.85
     unden
    0.84
     dilig
    0.83
     NOTICE
    0.75
    etheus
    0.72
    chwitz
    0.72
    puter
    0.71
    sterdam
    0.70
    iciary
    0.70
    uthor
    0.70
    Activations Density 0.006%

    No Known Activations