INDEX

Explanations

information related to allegations of abuse and misconduct

oai_token-act-pair · gpt-3.5-turbo

references to allegations or claims of abuse or exploitation

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Alright

-0.71

 Tycoon

-0.70

worthiness

-0.69

})

-0.69

TBD

-0.68

 resides

-0.66

Balance

-0.64

focus

-0.63

 optimize

-0.63

bye

-0.63

POSITIVE LOGITS

 harassed

1.00

 witnessing

0.93

 raped

0.92

 mist

0.90

 bullied

0.90

 discriminated

0.89

 coerced

0.88

 humiliated

0.85

 grop

0.84

 subjected

0.83

Activations Density 0.396%

information related to allegations of abuse and misconduct

references to allegations or claims of abuse or exploitation

No Comments

No Known Activations