INDEX

Explanations

references to the importance or significance of specific roles in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of the concept of "role" 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 @ 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

Latest

-0.71

Uri

-0.70

 Gleaming

-0.69

Interstitial

-0.69

bones

-0.63

Dream

-0.62

mort

-0.62

 sight

-0.62

Lev

-0.60

Pear

-0.60

POSITIVE LOGITS

playing

1.01

 reversal

0.88

 role

0.83

 roles

0.81

role

0.77

iotic

0.73

uties

0.73

 Role

0.72

iotics

0.72

inic

0.71

Activations Density 0.037%

references to the importance or significance of specific roles in various contexts

mentions of the concept of "role" in various contexts

No Comments

No Known Activations