INDEX

Explanations

mentions of retired military or law enforcement personnel

oai_token-act-pair · gpt-3.5-turbo

references to retired individuals, particularly those with military or authoritative backgrounds

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

otion

-0.74

 Grimm

-0.73

orie

-0.70

abal

-0.68

oos

-0.67

owitz

-0.67

insk

-0.66

Ã±

-0.66

bender

-0.65

belt

-0.65

POSITIVE LOGITS

 colonel

0.89

 retiring

0.82

oslov

0.77

 obsolete

0.76

 Marshal

0.76

 retire

0.74

 Colonel

0.73

dinand

0.72

 pensions

0.72

 soldier

0.72

Activations Density 0.032%

mentions of retired military or law enforcement personnel

references to retired individuals, particularly those with military or authoritative backgrounds

No Comments

No Known Activations