INDEX

Explanations

keywords related to the act of firing or termination of employment

oai_token-act-pair · gpt-3.5-turbo

instances of the word "fired" and its variations related to dismissal or termination

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

 amen

-0.74

 Towns

-0.70

 Revival

-0.64

sd

-0.62

flix

-0.62

lihood

-0.61

bitcoin

-0.60

sson

-0.60

 Dreams

-0.60

aque

-0.60

POSITIVE LOGITS

mails

0.92

bats

0.89

balls

0.84

uracy

0.80

bomb

0.79

 Fired

0.77

 salute

0.77

lessly

0.77

ragon

0.76

locks

0.76

Activations Density 0.047%

keywords related to the act of firing or termination of employment

instances of the word "fired" and its variations related to dismissal or termination

No Comments

No Known Activations