INDEX

Explanations

terminology related to sexual offenses and misconduct

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mdl

-0.67

 Certo

-0.64

nadequate

-0.63

anthene

-0.62

toplasmic

-0.62

 valle

-0.62

 Verdi

-0.61

RenderAtEndOf

-0.61

irical

-0.61

приятия

-0.60

POSITIVE LOGITS

 sexual

1.93

sex

1.89

SEX

1.85

 Sexual

1.83

Sex

1.82

SEX

1.76

Sexual

1.75

Sex

1.73

 sexuelle

1.63

sexual

1.60

Activations Density 0.063%

terminology related to sexual offenses and misconduct

No Comments

No Known Activations