INDEX

Explanations

the concept of harmful or violent cycles or behaviors

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "vicious" cycles or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ITNESS

-0.89

ovember

-0.86

inoa

-0.82

mberg

-0.81

brance

-0.81

ucket

-0.80

ablished

-0.78

ĸļ

-0.77

inez

-0.77

ourced

-0.77

POSITIVE LOGITS

ly

1.02

 predators

0.84

 venom

0.84

 vicious

0.83

 circle

0.81

 retribution

0.80

icious

0.79

abuse

0.77

 predator

0.75

 bite

0.74

Activations Density 0.048%

the concept of harmful or violent cycles or behaviors

references to the concept of "vicious" cycles or actions

No Comments

No Known Activations