INDEX

Explanations

words related to the action of speaking or pronouncing

oai_token-act-pair · gpt-3.5-turbo

references to oral arguments and discussions

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

vell

-0.80

etter

-0.79

IELD

-0.73

 guiActiveUnfocused

-0.72

ICS

-0.70

ELF

-0.69

lower

-0.69

ukong

-0.69

ERAL

-0.69

âĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪ

-0.67

POSITIVE LOGITS

 cavity

1.09

 oral

0.92

 contraceptive

0.91

 contraceptives

0.89

 ingestion

0.88

 hygiene

0.81

muc

0.79

 polio

0.78

 examination

0.77

 examinations

0.74

Activations Density 0.016%

words related to the action of speaking or pronouncing

references to oral arguments and discussions

No Comments

No Known Activations