INDEX

Explanations

terms related to fantasy RPG character classes

oai_token-act-pair · gpt-3.5-turbo

names of character classes from fantasy role-playing games

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

ĸļ

-0.79

idays

-0.76

iqueness

-0.76

oho

-0.76

ilion

-0.73

pty

-0.67

raltar

-0.66

irsch

-0.64

MSG

-0.64

itional

-0.63

POSITIVE LOGITS

llan

1.26

 Templ

0.81

 Corps

0.79

bane

0.78

 Geral

0.77

 Swordsman

0.77

 Colleges

0.74

 spirits

0.74

rior

0.73

 corps

0.73

Activations Density 0.096%

terms related to fantasy RPG character classes

names of character classes from fantasy role-playing games

No Comments

No Known Activations