INDEX

Explanations

verbs related to interaction or involvement

oai_token-act-pair · gpt-3.5-turbo

instances of the word "engage" and its variations, indicating a focus on interactions and involvement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

pled

-0.73

ple

-0.72

aver

-0.68

patient

-0.67

ichael

-0.66

ocrats

-0.64

ark

-0.63

ources

-0.62

Origin

-0.60

arily

-0.60

POSITIVE LOGITS

 engaged

0.94

 engagement

0.91

 engage

0.86

 engages

0.84

glomer

0.82

encer

0.78

TAIN

0.76

lear

0.75

lished

0.74

ments

0.74

Activations Density 0.012%

verbs related to interaction or involvement

instances of the word "engage" and its variations, indicating a focus on interactions and involvement

No Comments

No Known Activations