INDEX

Explanations

phrases related to comparisons or relationships between different entities or attributes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating strength or prioritization

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

iest

-0.69

inois

-0.69

utherland

-0.68

icter

-0.65

efully

-0.64

olulu

-0.62

ulla

-0.62

acci

-0.62

ertodd

-0.62

igmatic

-0.61

POSITIVE LOGITS

 sorts

0.78

 course

0.74

ADVERTISEMENT

0.67

·

0.65

=/

0.64

ELD

0.63

 Excellence

0.63

¾

0.61

ĵ

0.61

Life

0.60

Activations Density 0.307%

phrases related to comparisons or relationships between different entities or attributes

phrases indicating strength or prioritization

No Comments

No Known Activations