INDEX

Explanations

words related to encryption and cybersecurity

oai_token-act-pair · gpt-3.5-turbo

instances of the prefix "Ex," often indicating things related to exceptions or exceptional cases

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

HCR

-0.73

Kuh

-0.71

enegger

-0.70

 Learns

-0.68

LINE

-0.68

 Carib

-0.67

arity

-0.66

externalActionCode

-0.64

OUP

-0.64

 Sawyer

-0.63

POSITIVE LOGITS

haust

1.18

terior

1.13

ternally

0.99

cerpt

0.98

uber

0.97

changes

0.97

isting

0.92

posed

0.88

cessive

0.88

erc

0.88

Activations Density 0.007%

words related to encryption and cybersecurity

instances of the prefix "Ex," often indicating things related to exceptions or exceptional cases

No Comments

No Known Activations