INDEX

Explanations

phrases related to negative attributes or conditions

oai_token-act-pair · gpt-3.5-turbo

terms related to negative situations associated with illness or problems

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

ulhu

-0.80

Zup

-0.75

 antioxid

-0.73

Valid

-0.72

Factor

-0.70

illon

-0.69

ħĭ

-0.69

 Enhancement

-0.69

 è£ıè

-0.68

 Cosponsors

-0.67

POSITIVE LOGITS

gotten

0.93

nesses

0.69

 maiden

0.69

utton

0.66

iquid

0.65

phia

0.64

 misc

0.64

ady

0.64

Beck

0.63

 Lakers

0.63

Activations Density 0.050%

phrases related to negative attributes or conditions

terms related to negative situations associated with illness or problems

No Comments

No Known Activations