© 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. 20269
    Prev
    Next
    INDEX
    Explanations

    emotionally charged and impactful words or phrases

    oai_token-act-pair · gpt-3.5-turbo

    abstract concepts related to morality, conflict, and the human experience

    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
    CHAT
    -0.78
    ificant
    -0.74
    ificantly
    -0.66
    zl
    -0.62
    APD
    -0.62
    azeera
    -0.61
    eatures
    -0.60
    ittees
    -0.59
    ersen
    -0.59
    oops
    -0.58
    POSITIVE LOGITS
    ankind
    0.85
    lessness
    0.78
     emanating
    0.68
    thood
    0.67
    fulness
    0.66
     fame
    0.65
     itself
    0.65
    nesia
    0.65
    beard
    0.64
    ropy
    0.63
    Activations Density 0.466%

    No Known Activations