INDEX

Explanations

mentions of visits or visits to different places

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "visit" and its variants

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

claimer

-0.80

dp

-0.74

direction

-0.73

frac

-0.69

catentry

-0.67

xual

-0.67

 nutshell

-0.67

WAY

-0.67

communications

-0.67

ubi

-0.65

POSITIVE LOGITS

 museums

1.01

 Disneyland

0.93

 Hiroshima

0.88

 relatives

0.86

 temples

0.86

 Auschwitz

0.84

 graves

0.81

 abroad

0.81

 villages

0.81

 campuses

0.80

Activations Density 0.075%

mentions of visits or visits to different places

occurrences of the word "visit" and its variants

No Comments

No Known Activations