INDEX

Explanations

phrases related to committing crimes or unknowingly being involved in criminal activities

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

 arroyo

-0.53

ffiche

-0.49

 minera

-0.47

LayoutStyle

-0.46

 revan

-0.45

 incidente

-0.43

 alpina

-0.42

 barran

-0.42

 oliva

-0.42

 pilar

-0.41

POSITIVE LOGITS

 ourselves

1.47

our

0.94

we

0.88

our

0.84

Our

0.82

OUR

0.82

 ours

0.82

We

0.81

we

0.81

Our

0.80

Activations Density 0.740%

phrases related to committing crimes or unknowingly being involved in criminal activities

No Comments

No Known Activations