INDEX

Explanations

keywords related to conjunctions and transitional phrases

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "And."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ãģĨ

-0.80

sense

-0.72

cale

-0.66

fell

-0.65

xes

-0.65

manship

-0.64

dayName

-0.64

sic

-0.63

ciples

-0.62

viation

-0.62

POSITIVE LOGITS

romeda

1.54

rea

1.34

hra

0.93

ersen

0.89

alus

0.87

rogens

0.86

rew

0.82

vo

0.79

withstanding

0.78

rey

0.77

Activations Density 0.058%

keywords related to conjunctions and transitional phrases

occurrences of the word "And."

No Comments

No Known Activations