INDEX

Explanations

phrases describing various situations or occurrences that involve multiple affected individuals

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases emphasizing quantity and existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Shutterstock

-0.85

ĸļ

-0.77

 Ferry

-0.74

 Vulcan

-0.72

 Donkey

-0.72

 Aging

-0.70

 Consent

-0.67

 Selection

-0.65

 Alone

-0.64

 Athletics

-0.64

POSITIVE LOGITS

ilers

0.85

ensical

0.83

ahs

0.78

bang

0.77

enos

0.77

doms

0.76

 besides

0.75

 alike

0.75

related

0.74

 imaginable

0.73

Activations Density 0.664%

phrases describing various situations or occurrences that involve multiple affected individuals

repetitive phrases emphasizing quantity and existence

No Comments

No Known Activations