INDEX

Explanations

phrases related to capabilities, skills, or aptitude

oai_token-act-pair · gpt-3.5-turbo

mentions of "ability" or references to capabilities and competencies

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ummies

-0.67

quished

-0.64

mpire

-0.62

ogi

-0.61

âĿ

-0.61

Reply

-0.61

lake

-0.60

busters

-0.59

TAG

-0.59

lie

-0.59

POSITIVE LOGITS

 ability

3.61

 abilities

2.34

 Ability

2.23

 capability

2.04

 inability

2.02

Ability

1.81

 capacity

1.67

 willingness

1.65

 capabilities

1.61

 Abilities

1.58

Activations Density 0.023%

phrases related to capabilities, skills, or aptitude

mentions of "ability" or references to capabilities and competencies

No Comments

No Known Activations