INDEX

Explanations

phrases indicating strong positive appreciation or approval

oai_token-act-pair · gpt-3.5-turbo

expressions of significant appreciation or benefit

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

esome

-0.71

 Notting

-0.71

aple

-0.70

ulia

-0.70

smoking

-0.69

 Tags

-0.69

nar

-0.66

ties

-0.66

onomy

-0.65

tein

-0.65

POSITIVE LOGITS

 appreciated

0.94

 benefited

0.94

 appreci

0.85

 appre

0.84

 amounts

0.79

 depends

0.77

 appreciate

0.76

 impacting

0.75

 benef

0.73

 effected

0.73

Activations Density 0.017%

phrases indicating strong positive appreciation or approval

expressions of significant appreciation or benefit

No Comments

No Known Activations