INDEX

Explanations

mentions of wars or conflicts

oai_token-act-pair · gpt-3.5-turbo

references to warfare and conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

essee

-0.83

AME

-0.67

 Investigator

-0.63

 Universe

-0.63

 drawer

-0.58

UTE

-0.58

ECT

-0.57

Aad

-0.57

 Seller

-0.56

 Collider

-0.56

POSITIVE LOGITS

lords

1.30

lord

1.23

ring

1.14

riors

1.12

planes

1.12

ping

1.11

rior

1.06

fare

1.04

zone

1.02

fighter

1.02

Activations Density 0.048%

mentions of wars or conflicts

references to warfare and conflict

No Comments

No Known Activations