INDEX

Explanations

personality-related terms and traits

oai_token-act-pair · gpt-3.5-turbo

references to personality traits

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

LOAD

-0.74

ells

-0.73

anus

-0.71

WARE

-0.70

een

-0.70

ERY

-0.69

scl

-0.69

yer

-0.69

eries

-0.68

REAM

-0.67

POSITIVE LOGITS

 trait

0.97

 disorder

0.97

 personality

0.97

istically

0.96

 traits

0.95

istics

0.85

 charisma

0.82

istical

0.80

 quirks

0.79

 personalities

0.79

Activations Density 0.019%

personality-related terms and traits

references to personality traits

No Comments

No Known Activations