INDEX

Explanations

instances where actions are done in an autonomous or self-sufficient manner

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the concept of independence

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

mers

-0.79

ger

-0.68

 Millennium

-0.67

Ing

-0.67

ties

-0.66

GER

-0.65

Nin

-0.63

 Nieto

-0.63

mer

-0.63

Zar

-0.62

POSITIVE LOGITS

 corrobor

0.81

 owned

0.80

 identifiable

0.78

wine

0.78

 verified

0.77

 minded

0.75

etitive

0.74

 authenticated

0.72

 detonated

0.72

 discl

0.71

Activations Density 0.010%

instances where actions are done in an autonomous or self-sufficient manner

repeated mentions of the concept of independence

No Comments

No Known Activations