INDEX

Explanations

words related to objections, conflicts, and negotiations

oai_token-act-pair · gpt-3.5-turbo

situations involving strong emotions or appeals for help

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

olars

-0.75

utterstock

-0.74

urches

-0.70

vernment

-0.69

utical

-0.69

ioxid

-0.68

 APPLIC

-0.66

cano

-0.66

arent

-0.66

rontal

-0.66

POSITIVE LOGITS

 himself

1.32

his

1.22

him

1.07

 hers

0.82

 Schwarzenegger

0.77

 TAMADRA

0.75

his

0.73

 Meier

0.73

 Gamble

0.71

he

0.71

Activations Density 0.725%

words related to objections, conflicts, and negotiations

situations involving strong emotions or appeals for help

No Comments

No Known Activations