INDEX

Explanations

words related to mental health or behavior

oai_token-act-pair · gpt-3.5-turbo

references to madness or lunacy

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

LECT

-0.69

 Proced

-0.68

åĬ

-0.67

ĵĺ

-0.63

ortment

-0.62

redo

-0.61

Lot

-0.60

 Gamble

-0.60

 Dome

-0.60

âĢ¢âĢ¢

-0.57

POSITIVE LOGITS

acy

1.38

atic

1.33

atics

1.29

acies

1.27

cheon

1.26

atical

1.18

etic

1.12

acists

1.09

atural

1.07

aries

1.04

Activations Density 0.055%

words related to mental health or behavior

references to madness or lunacy

No Comments

No Known Activations