INDEX

Explanations

the word "notice" across various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "notice" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

quer

-0.74

ãĥ³ãĤ¸

-0.70

 rall

-0.68

ccording

-0.67

iera

-0.66

ãĥĥãĥī

-0.65

 ancest

-0.63

tre

-0.63

 weap

-0.62

 cheek

-0.62

POSITIVE LOGITS

 notice

1.02

 Notice

0.95

 notices

0.87

ably

0.83

andum

0.81

lessly

0.80

 Signs

0.79

notice

0.77

lication

0.77

Notice

0.76

Activations Density 0.008%

the word "notice" across various contexts

instances of the word "notice" in various contexts

No Comments

No Known Activations