INDEX

Explanations

phrases related to advertising and marketing strategies

oai_token-act-pair · gpt-3.5-turbo

ellipses indicating omissions or pauses in text

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

 mobilized

-0.71

 outl

-0.70

 greens

-0.69

 merit

-0.69

 strength

-0.68

 depletion

-0.68

 plur

-0.67

 advoc

-0.66

 empath

-0.66

 strokes

-0.66

POSITIVE LOGITS

BUT

1.06

sort

0.94

they

0.91

why

0.90

oops

0.88

interesting

0.87

 âĢİ

0.87

until

0.85

yet

0.84

wow

0.83

Activations Density 0.035%

phrases related to advertising and marketing strategies

ellipses indicating omissions or pauses in text

No Comments

No Known Activations