INDEX

Explanations

phrases indicating willingness or acceptance

oai_token-act-pair · gpt-3.5-turbo

positive emotional states or expressions of happiness

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 disproportion

-0.73

 GOODMAN

-0.68

oons

-0.67

Posts

-0.66

senal

-0.64

hur

-0.64

 Relief

-0.64

arnaev

-0.63

ulz

-0.61

anan

-0.61

POSITIVE LOGITS

 embraced

0.88

 awaiting

0.81

 accepted

0.80

 awaited

0.78

 complied

0.78

 transitioned

0.77

 supplied

0.76

 acknowledged

0.76

 parted

0.75

 welcomed

0.75

Activations Density 0.097%

phrases indicating willingness or acceptance

positive emotional states or expressions of happiness

No Comments

No Known Activations