INDEX

Explanations

words related to being inexperienced or simplistic in their understanding

oai_token-act-pair · gpt-3.5-turbo

the term "naive" and its variations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Downloadha

-0.79

ŃĶ

-0.73

interrupted

-0.71

 adjoining

-0.70

inders

-0.66

hops

-0.66

 Occupations

-0.66

orset

-0.65

 antidepressant

-0.65

 hotly

-0.65

POSITIVE LOGITS

 naive

1.28

 naÃ¯ve

1.12

Ã¯ve

0.87

sters

0.86

glers

0.83

poke

0.78

 innocence

0.76

cipled

0.76

fusc

0.74

lings

0.74

Activations Density 0.007%

words related to being inexperienced or simplistic in their understanding

the term "naive" and its variations in various contexts

No Comments

No Known Activations