INDEX

Explanations

phrases indicating recommendations or instructions

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases indicating a need to monitor or keep track of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ALLY

-0.85

bernatorial

-0.78

ĪĴ

-0.75

 PRESIDENT

-0.75

BAT

-0.69

®

-0.68

 behalf

-0.65

haps

-0.64

«

-0.64

ISM

-0.64

POSITIVE LOGITS

 afloat

0.94

 tabs

0.82

 footing

0.78

awed

0.72

 steady

0.72

 wraps

0.71

 composure

0.67

lid

0.66

eye

0.66

iping

0.65

Activations Density 0.123%

phrases indicating recommendations or instructions

repetitive phrases indicating a need to monitor or keep track of something

No Comments

No Known Activations