INDEX

Explanations

instances of the word "detect" along with related phrases like "detect and treat"

oai_token-act-pair · gpt-3.5-turbo

terms related to the act of identifying or discovering something

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asin

-0.85

skill

-0.82

perty

-0.82

mort

-0.79

yss

-0.76

 Vide

-0.73

atern

-0.72

esa

-0.72

par

-0.70

sl

-0.70

POSITIVE LOGITS

 detect

1.30

 detecting

0.99

 detects

0.98

 detected

0.98

 detection

0.96

 detectable

0.92

 Detection

0.91

 Detected

0.81

 Detect

0.80

ected

0.74

Activations Density 0.015%

instances of the word "detect" along with related phrases like "detect and treat"

terms related to the act of identifying or discovering something

No Comments

No Known Activations