INDEX

Explanations

names of characters from movies or fictional accounts, especially when someone leaves or is involved in a friendship

oai_token-act-pair · gpt-3.5-turbo

elements related to fictional narratives and character relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Advertisements

-0.89

undy

-0.77

ascript

-0.73

âĹ¼

-0.71

HL

-0.67

imer

-0.67

rans

-0.67

ariat

-0.66

catentry

-0.64

Insert

-0.64

POSITIVE LOGITS

 confronts

0.61

 Seeking

0.59

adiator

0.59

 learns

0.58

 fugitive

0.57

Ĭ±

0.56

Adds

0.56

ixtape

0.55

 Makoto

0.55

 tensions

0.55

Activations Density 0.062%

names of characters from movies or fictional accounts, especially when someone leaves or is involved in a friendship

elements related to fictional narratives and character relationships

No Comments

No Known Activations