INDEX

Explanations

phrases related to different types of roles or assignments

oai_token-act-pair · gpt-3.5-turbo

terms related to roles and role-playing 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

False

-0.69

 Bulg

-0.68

bow

-0.64

incinn

-0.63

REAM

-0.63

yg

-0.62

Oo

-0.62

Latest

-0.62

undred

-0.60

bones

-0.60

POSITIVE LOGITS

playing

1.09

 reversal

0.91

 roles

0.88

players

0.77

played

0.76

entials

0.74

 models

0.73

model

0.73

plays

0.72

 vacated

0.71

Activations Density 0.041%

phrases related to different types of roles or assignments

terms related to roles and role-playing in various contexts

No Comments

No Known Activations