INDEX

Explanations

superlatives indicating extreme qualities or accomplishments

oai_token-act-pair · gpt-3.5-turbo

phrases indicating rankings or classifications of items based on their significance or quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

bleacher

-0.65

pmwiki

-0.60

gow

-0.58

Fill

-0.57

 Unsure

-0.57

EStream

-0.56

othal

-0.55

GU

-0.55

umeric

-0.54

ctic

-0.54

POSITIVE LOGITS

any

0.89

 theirs

0.88

all

0.86

 hers

0.84

0.83

0.81

0.80

0.78

 them

0.74

Activations Density 0.101%

superlatives indicating extreme qualities or accomplishments

phrases indicating rankings or classifications of items based on their significance or quality

No Comments

No Known Activations