INDEX

Explanations

mentions of military and reserve-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to military service and ranks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

theless

-0.85

ulent

-0.77

FH

-0.74

Elsa

-0.73

 Collider

-0.71

ulence

-0.71

Property

-0.71

Shares

-0.70

properties

-0.68

quickShipAvailable

-0.67

POSITIVE LOGITS

 colonel

1.11

 stationed

1.09

 Corps

1.04

 Infantry

1.03

 fatig

1.01

 Commando

1.00

 WWII

0.98

geant

0.98

 Marines

0.97

 platoon

0.96

Activations Density 0.238%

mentions of military and reserve-related terms

terms related to military service and ranks

No Comments

No Known Activations