INDEX

Explanations

the word "amp" and its variants, indicating a focus on amplification or intensity 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 @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

SHIP

-0.96

BLIC

-0.78

ģĸ

-0.75

 Kers

-0.73

ignty

-0.69

ĪĴ

-0.65

 resemblance

-0.65

ryu

-0.62

ALLY

-0.61

bis

-0.61

POSITIVE LOGITS

aign

1.14

agne

1.09

hetamine

1.07

aic

1.03

oline

1.00

shire

0.98

eters

0.96

stead

0.94

owered

0.92

agna

0.89

Activations Density 0.027%

the word "amp" and its variants, indicating a focus on amplification or intensity in various contexts

No Comments

No Known Activations