INDEX

Explanations

rules or instructions related to character stats or abilities in a role-playing game

oai_token-act-pair · gpt-3.5-turbo

terms related to character attributes and progression in a role-playing game context

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

-0.80

çļ

-0.75

DJ

-0.72

ettel

-0.70

Ò

-0.69

à¼

-0.68

ãĥ¼ãĥĨãĤ£

-0.68

LH

-0.67

market

-0.67

acion

-0.67

POSITIVE LOGITS

 Arcane

1.07

 Dexterity

1.05

nec

1.04

 sorcerer

1.01

 Charisma

0.99

 cleric

0.98

 arcane

0.90

 Spell

0.90

 clerics

0.90

spell

0.89

Activations Density 0.151%

rules or instructions related to character stats or abilities in a role-playing game

terms related to character attributes and progression in a role-playing game context

No Comments

No Known Activations