INDEX

Explanations

instances where someone is directly involved in an action or decision

oai_token-act-pair · gpt-3.5-turbo

instances of emotional reactions and interactions among characters

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ãĥĺãĥ©

-0.73

 eleph

-0.67

ammy

-0.63

translation

-0.62

 Banana

-0.60

lat

-0.59

 thous

-0.59

 Collider

-0.59

inia

-0.59

ictionary

-0.57

POSITIVE LOGITS

hement

0.91

 urge

0.91

 immediately

0.87

 empath

0.86

 instinctively

0.85

 hesitate

0.84

 empathy

0.78

 accordingly

0.76

 promptly

0.76

 compassion

0.75

Activations Density 0.837%

instances where someone is directly involved in an action or decision

instances of emotional reactions and interactions among characters

No Comments

No Known Activations