© 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. 4-RES-JB
    11. 11005
    Prev
    Next
    INDEX
    Explanations

    the verb "constitute" used in relation to defining or making up something

    oai_token-act-pair · gpt-3.5-turbo

    the word "constitute" and its variations, indicating discussions about definitions or classifications of actions or conditions

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

    No Comments

    Negative Logits
     Bald
    -0.69
    ying
    -0.66
    hner
    -0.66
    ithing
    -0.62
    ITCH
    -0.62
    enger
    -0.61
    itch
    -0.59
    hon
    -0.59
    hy
    -0.58
     preacher
    -0.57
    POSITIVE LOGITS
    aldehyde
    0.84
    idable
    0.83
    chnology
    0.79
    itutes
    0.76
    utive
    0.75
    ances
    0.71
    ally
    0.71
    lehem
    0.71
    amaz
    0.70
    ALLY
    0.69
    Activations Density 0.020%

    No Known Activations