INDEX

Explanations

phrases related to criminal activities and investigations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 cushi

-1.12

 swarovski

-1.12

 tupperware

-1.11

 peppa

-1.10

 Ename

-1.09

 affor

-1.09

 oreo

-1.04

 murano

-1.04

 nutella

-1.04

 eiffel

-1.04

POSITIVE LOGITS

 inquest

0.75

 autopsy

0.68

 investigation

0.67

 investigations

0.65

 kriminal

0.63

İstinadlar

0.62

 investigating

0.61

 investiga

0.60

InitVars

0.59

investigation

0.59

Activations Density 0.573%

phrases related to criminal activities and investigations

No Comments

No Known Activations