INDEX

Explanations

references to pornographic material and related content

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Sklici

-0.61

醐

-0.57

 solidar

-0.55

Reparto

-0.54

Referencie

-0.52

Kör

-0.52

Kö

-0.49

Zunanje

-0.49

Merk

-0.48

 kapital

-0.48

POSITIVE LOGITS

 porn

1.23

 pornography

1.11

 Porn

1.10

Porn

0.95

porn

0.94

 pamph

0.71

 unspeak

0.65

 surpl

0.63

 subgoals

0.62

 bourg

0.62

Activations Density 0.071%

references to pornographic material and related content

No Comments

No Known Activations