INDEX

Explanations

the word "spike" at different levels of intensity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "spike" related to increases or surges in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ITNESS

-0.70

vironments

-0.68

İ

-0.67

 Incarnation

-0.63

ĨĴ

-0.62

voc

-0.62

YA

-0.62

Pil

-0.62

velop

-0.61

________________________________

-0.61

POSITIVE LOGITS

 spike

1.23

 spikes

1.15

olicy

0.85

rings

0.79

balls

0.72

ãĥ©ãĥ³

0.72

 bracelet

0.70

 spiked

0.70

 snowball

0.65

acers

0.65

Activations Density 0.007%

the word "spike" at different levels of intensity

instances of the word "spike" related to increases or surges in various contexts

No Comments

No Known Activations