INDEX

Explanations

specific mentions of territories or territories that are being discussed in relation to different topics or events

oai_token-act-pair · gpt-3.5-turbo

references to geographical or political territories

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

odcast

-0.86

nder

-0.71

kus

-0.70

artz

-0.70

uth

-0.68

uster

-0.67

Bi

-0.66

racted

-0.65

instein

-0.65

strument

-0.64

POSITIVE LOGITS

 boundaries

0.90

 conquered

0.88

 Territories

0.83

 territories

0.82

 inhabited

0.81

 annexed

0.79

 bordering

0.77

 territory

0.75

 adjoining

0.73

 unpop

0.73

Activations Density 0.029%

specific mentions of territories or territories that are being discussed in relation to different topics or events

references to geographical or political territories

No Comments

No Known Activations