INDEX

Explanations

personal pronouns associated with specific actions or qualities

oai_token-act-pair · gpt-3.5-turbo

pronouns referring to the reader or speaker

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

hap

-0.67

 Wake

-0.64

Hok

-0.59

 obstruction

-0.57

ãĤ¨

-0.57

 endif

-0.57

 Seah

-0.57

 Ballard

-0.56

 Sund

-0.55

detail

-0.55

POSITIVE LOGITS

 wont

0.75

'll

0.73

cius

0.71

izons

0.69

 surely

0.66

olit

0.66

erd

0.66

taboola

0.65

pton

0.65

sembly

0.65

Activations Density 0.179%

personal pronouns associated with specific actions or qualities

pronouns referring to the reader or speaker

No Comments

No Known Activations