INDEX

Explanations

phrases related to providing evidence or proof

oai_token-act-pair · gpt-3.5-turbo

phrases related to proving or confirming truths or assumptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

soever

-0.71

 Shutterstock

-0.68

 Springer

-0.67

ĸļ

-0.66

 undone

-0.66

 paced

-0.63

lda

-0.62

 alone

-0.62

 ILCS

-0.61

yip

-0.61

POSITIVE LOGITS

 existence

1.47

 authenticity

1.38

 validity

1.32

worthiness

1.17

existence

1.16

 innocence

1.13

iability

1.09

 legitimacy

1.08

 correctness

1.07

 accuracy

1.02

Activations Density 0.277%

phrases related to providing evidence or proof

phrases related to proving or confirming truths or assumptions

No Comments

No Known Activations