INDEX

Explanations

phrases related to locking or being locked up

oai_token-act-pair · gpt-3.5-turbo

instances of the word "locked" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

resso

-0.75

nor

-0.73

annis

-0.73

lus

-0.71

issance

-0.69

arenthood

-0.68

conom

-0.68

anas

-0.67

anners

-0.64

ãĤ¡

-0.64

POSITIVE LOGITS

 locked

1.05

 locks

0.84

 lock

0.83

locked

0.82

 recess

0.80

locking

0.78

picking

0.78

door

0.78

heed

0.76

 doors

0.72

Activations Density 0.007%

phrases related to locking or being locked up

instances of the word "locked" in various contexts

No Comments

No Known Activations