INDEX

Explanations

information at the beginning and end of a news article or webpage

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

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.6.hook_resid_post_16384_anthropic_fast_lr

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.6.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

 massacres

-0.92

corruption

-0.91

ABORT

-0.89

ायर

-0.87

 لينك

-0.87

 disinformation

-0.86

 bombings

-0.85

 atrocities

-0.85

ALREADY

-0.84

 insurgents

-0.82

POSITIVE LOGITS

<bos>

16.41

expandindo

2.63

GEBURTSDATUM

2.61

 dispen

2.59

 encomp

2.57

fuf

2.49

 betweenstory

2.48

 guarante

2.47

 affor

2.47

 accla

2.44

Activations Density 0.011%

information at the beginning and end of a news article or webpage

No Comments

No Known Activations