INDEX

Explanations

linguistic tokens denoted by special characters or numbers, possibly used for encoding information

oai_token-act-pair · gpt-3.5-turbo

quantifiable risks and issues associated with safety and health

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

 endeavour

-0.61

tics

-0.57

Tro

-0.57

 exerc

-0.57

scill

-0.56

 assignment

-0.55

 charm

-0.54

Ń·

-0.54

atility

-0.54

 ideal

-0.53

POSITIVE LOGITS

Advertisement

1.08

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

1.05

ÂłÂłÂłÂł

1.05

advertisement

1.02

Newsletter

1.01

ÂłÂłÂłÂłÂłÂłÂłÂł

0.97

posted

0.90

Article

0.86

ÂłÂłÂł

0.83

AMY

0.83

Activations Density 0.627%

linguistic tokens denoted by special characters or numbers, possibly used for encoding information

quantifiable risks and issues associated with safety and health

No Comments

No Known Activations