INDEX

Explanations

verbs associated with persuasion or convincing

oai_token-act-pair · gpt-3.5-turbo

instances of the word "convince" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pmwiki

-0.70

 Miko

-0.70

alm

-0.67

 Rover

-0.64

OIL

-0.64

 flourishing

-0.63

 Occupations

-0.63

estial

-0.62

 allotted

-0.60

abytes

-0.60

POSITIVE LOGITS

ingly

0.86

 convince

0.86

tale

0.86

ments

0.84

 convinc

0.81

 convinced

0.78

-+-+

0.74

 persuade

0.73

imaru

0.72

 persuaded

0.72

Activations Density 0.017%

verbs associated with persuasion or convincing

instances of the word "convince" in various contexts

No Comments

No Known Activations