INDEX

Explanations

phrases related to direct actions or directives

oai_token-act-pair · gpt-3.5-turbo

the term "Direct" and its variations, indicating a focus on direct content or services

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

OAD

-0.86

 Schne

-0.77

anners

-0.71

 Haram

-0.71

FORE

-0.69

EAR

-0.68

glers

-0.68

 Kaufman

-0.66

 Vance

-0.65

OPLE

-0.65

POSITIVE LOGITS

ions

1.04

iate

1.02

ives

0.95

edly

0.92

ories

0.89

eur

0.84

enture

0.83

lined

0.82

irect

0.80

ing

0.80

Activations Density 0.018%

phrases related to direct actions or directives

the term "Direct" and its variations, indicating a focus on direct content or services

No Comments

No Known Activations