INDEX

Explanations

adjectives related to qualities or characteristics

oai_token-act-pair · gpt-3.5-turbo

phrases that describe types or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

UF

-0.86

dn

-0.83

 âĶľ

-0.82

ï¸

-0.82

none

-0.79

arat

-0.79

arest

-0.78

umber

-0.78

idelines

-0.78

APS

-0.76

POSITIVE LOGITS

 thing

1.57

 stuff

1.14

 shenanigans

1.08

 mischief

1.06

 behaviour

1.06

 behavior

1.05

 crap

1.04

 nonsense

1.02

 relationship

1.00

 situation

0.99

Activations Density 0.051%

adjectives related to qualities or characteristics

phrases that describe types or categories

No Comments

No Known Activations