© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 9-RES-JB
    11. 10367
    Prev
    Next
    INDEX
    Explanations

    phrases related to negative or harmful actions or behaviors, especially involving violence or illegal activities

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to abusive or predatory behavior

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     univers
    -0.76
    uitive
    -0.72
    ahime
    -0.71
     Revision
    -0.68
    usterity
    -0.68
    icult
    -0.67
    rastructure
    -0.67
     Effective
    -0.65
    ircraft
    -0.65
    ãĤ¤
    -0.64
    POSITIVE LOGITS
     stole
    1.09
     proceeded
    1.07
     drank
    1.05
     subsequently
    1.05
     ate
    1.03
     secondly
    1.03
     assaulted
    1.03
     raped
    1.02
     overheard
    1.02
     interfered
    1.02
    Activations Density 0.345%

    No Known Activations