INDEX

Explanations

phrases related to negative emotions, interpersonal conflict, and reactions

oai_token-act-pair · gpt-3.5-turbo

expressions of strong emotions or feelings, particularly negative ones

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

adoes

-0.61

 treaties

-0.60

 billions

-0.59

 schedule

-0.58

Thumbnail

-0.58

 Presidents

-0.55

 basics

-0.55

soDeliveryDate

-0.54

 monarchy

-0.54

 pioneered

-0.54

POSITIVE LOGITS

Another

1.14

 Another

1.07

another

1.02

 another

0.96

him

0.91

his

0.86

He

0.84

Someone

0.81

he

0.80

 Someone

0.76

Activations Density 0.480%

phrases related to negative emotions, interpersonal conflict, and reactions

expressions of strong emotions or feelings, particularly negative ones

No Comments

No Known Activations