INDEX

Explanations

 references to trust and related concepts

oai_token-act-pair · gpt-3.5-turbo

references to trust and trustworthiness in relationships or systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

pmwiki

-0.79

 Vide

-0.69

ople

-0.68

nesota

-0.63

ãĥ©ãĥ³

-0.62

 Gale

-0.62

Mek

-0.61

ploy

-0.60

 Despair

-0.60

OPLE

-0.59

POSITIVE LOGITS

worthiness

1.75

ees

1.36

worthy

1.27

ee

1.01

lessly

0.93

ful

0.90

fulness

0.88

full

0.88

0.86

fully

0.85

Activations Density 0.031%

&nbsp;references to trust and related concepts

references to trust and trustworthiness in relationships or systems

No Comments

No Known Activations

references to trust and related concepts