INDEX

Explanations

phrases related to taking action or making an effort

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate going out of one's way or making an effort

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

kefeller

-0.75

cius

-0.70

cious

-0.64

gemony

-0.64

uncture

-0.64

hiba

-0.63

iosity

-0.63

Card

-0.63

CRIPTION

-0.62

Cre

-0.61

POSITIVE LOGITS

fitted

0.92

ranged

0.90

stretched

0.82

wards

0.80

casts

0.73

smart

0.69

cast

0.67

landish

0.66

onda

0.65

house

0.65

Activations Density 0.032%

phrases related to taking action or making an effort

phrases that indicate going out of one's way or making an effort

No Comments

No Known Activations