INDEX

Explanations

references to financial aspects, such as fortune or income, specifically emphasizing negative instances like errors

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

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

HuggingFaceFW/fineweb

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

-0.61

 архивлан

-0.56

دانشنامهٔ

-0.52

 geha

-0.51

queryInterface

-0.50

hã

-0.50

 виді

-0.50

richTextPanel

-0.50

 защото

-0.49

afterEach

-0.49

POSITIVE LOGITS

 fortune

2.79

 Fortune

2.71

Fortune

2.39

fortune

2.23

 fortunes

2.08

fortunes

1.76

 fortuna

1.38

 FORT

1.30

 stockholm

1.16

 fortun

1.15

Activations Density 0.200%

references to financial aspects, such as fortune or income, specifically emphasizing negative instances like errors

No Comments

No Known Activations

references to financial aspects, such as fortune or income, specifically emphasizing negative instances like errors

No Comments

No Known Activations