INDEX

Explanations

instances of the word "role" and its context

oai_token-act-pair · gpt-3.5-turbo

references to 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 @ 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

Latest

-0.76

Uri

-0.65

 Bulg

-0.64

 Bened

-0.64

False

-0.63

Sab

-0.63

mares

-0.62

Hig

-0.62

ãĥ©ãĥ³

-0.62

Lev

-0.61

POSITIVE LOGITS

 roles

0.93

playing

0.90

 role

0.86

role

0.80

 reversal

0.79

uty

0.74

ioned

0.73

entials

0.72

ional

0.72

annel

0.71

Activations Density 0.028%

instances of the word "role" and its context

references to the concept of "role" in various contexts

No Comments

No Known Activations