INDEX

Explanations

words related to emotions or moods, particularly focusing on positive or uplifting feelings

oai_token-act-pair · gpt-3.5-turbo

references to emotional states or changes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

advertisement

-0.72

 Prosecut

-0.72

pta

-0.68

ces

-0.67

 WATCHED

-0.67

 Universities

-0.67

Ded

-0.66

PLIC

-0.64

chnology

-0.63

xton

-0.63

POSITIVE LOGITS

 mood

0.96

 swings

0.88

 Mood

0.84

iness

0.83

0.82

 Indigo

0.76

achine

0.75

iest

0.73

 comed

0.72

ily

0.72

Activations Density 0.021%

words related to emotions or moods, particularly focusing on positive or uplifting feelings

references to emotional states or changes

No Comments

No Known Activations