INDEX

Explanations

phrases related to accusations or suspicions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating accusations or suspicions of wrongdoing

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

chini

-0.76

orkshire

-0.74

rio

-0.72

ussion

-0.70

ulla

-0.69

gomery

-0.68

uilt

-0.66

urrent

-0.64

itta

-0.63

eas

-0.63

POSITIVE LOGITS

Tre

0.73

 Â¯

0.67

âĺħâĺħ

0.65

 Slov

0.60

 Seeking

0.58

Naz

0.58

 Django

0.58

 Bring

0.57

 accompanies

0.57

dding

0.57

Activations Density 0.402%

phrases related to accusations or suspicions

phrases indicating accusations or suspicions of wrongdoing

No Comments

No Known Activations