INDEX

Explanations

mentions of military or protective organizations

oai_token-act-pair · gpt-3.5-turbo

references to various military and national guard entities

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

Sov

-0.67

STEP

-0.67

jad

-0.67

doi

-0.65

piring

-0.65

ãĤ¼ãĤ¦ãĤ¹

-0.63

Hop

-0.61

 Craw

-0.61

ãĥ´

-0.61

change

-0.61

POSITIVE LOGITS

 Guard

0.97

 Corps

0.83

 stationed

0.82

naire

0.80

heed

0.80

 uniforms

0.80

 patrolling

0.78

 salute

0.77

 Guards

0.75

 reserv

0.74

Activations Density 0.014%

mentions of military or protective organizations

references to various military and national guard entities

No Comments

No Known Activations