INDEX

Explanations

pronouns followed by a verb

oai_token-act-pair · gpt-3.5-turbo

instances of the word "We" indicating collective or organizational statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 gratification

-0.70

cum

-0.66

 Publication

-0.64

 Eleven

-0.60

quo

-0.59

LSD

-0.58

 totality

-0.58

UD

-0.57

 PUBLIC

-0.56

 guiActiveUnfocused

-0.56

POSITIVE LOGITS

've

1.24

're

1.15

'll

1.10

ldon

1.08

bsite

1.06

bley

1.05

akening

0.99

alth

0.97

selves

0.95

ighed

0.94

Activations Density 0.223%

pronouns followed by a verb

instances of the word "We" indicating collective or organizational statements

No Comments

No Known Activations