INDEX

Explanations

mentions of the term "suicide."

oai_token-act-pair · gpt-3.5-turbo

the suffix "su," indicating a focus on words related to suicide or concerning mental health issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Gutenberg

-0.87

 Turks

-0.68

heads

-0.66

 directions

-0.65

head

-0.65

 torque

-0.65

 pounding

-0.62

 covering

-0.61

 Germans

-0.60

 tightening

-0.60

POSITIVE LOGITS

icide

1.32

arez

1.31

zanne

1.22

pper

1.22

icides

1.17

icidal

1.13

perm

1.11

pport

1.05

itable

1.02

ply

0.96

Activations Density 0.013%

mentions of the term "suicide."

the suffix "su," indicating a focus on words related to suicide or concerning mental health issues

No Comments

No Known Activations