INDEX

Explanations

the phrase "or" used in comparisons between quantities or options

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantities or ranges

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

tnc

-0.72

ĸļ

-0.72

ESE

-0.69

Wiki

-0.69

ãĤ©

-0.68

Cth

-0.67

ONSORED

-0.66

advertisement

-0.66

ç«

-0.65

gha

-0.64

POSITIVE LOGITS

 apiece

0.82

 gallon

0.81

undred

0.81

ottest

0.71

imore

0.71

 acres

0.71

-$

0.69

 hour

0.69

 dozen

0.68

 ounce

0.65

Activations Density 0.348%

the phrase "or" used in comparisons between quantities or options

phrases indicating quantities or ranges

No Comments

No Known Activations