INDEX

Explanations

mentions of past work or familiarity in a specific field

oai_token-act-pair · gpt-3.5-turbo

phrases or terms related to the level of experience

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

laws

-0.73

corn

-0.71

fam

-0.69

 Pengu

-0.67

law

-0.66

ificant

-0.65

bye

-0.65

ission

-0.65

oos

-0.64

anski

-0.64

POSITIVE LOGITS

 firsthand

0.83

 Experience

0.79

ually

0.75

ooters

0.75

 veter

0.75

ience

0.74

mingham

0.72

 abroad

0.72

 experience

0.72

ienced

0.71

Activations Density 0.030%

mentions of past work or familiarity in a specific field

phrases or terms related to the level of experience

No Comments

No Known Activations