INDEX

Explanations

phrases related to the word "occurrence"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Occ" and related forms, likely indicating a focus on the term "occidental" or similar terms

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

HCR

-0.92

EMENT

-0.69

BIL

-0.69

NESS

-0.69

RAW

-0.68

DON

-0.67

culosis

-0.67

DI

-0.67

 Hilton

-0.67

åĤ

-0.67

POSITIVE LOGITS

idental

1.20

urrence

1.09

urred

1.00

urring

0.99

occ

0.96

asionally

0.95

urs

0.94

asions

0.89

rontal

0.86

uria

0.85

Activations Density 0.023%

phrases related to the word "occurrence"

occurrences of the word "Occ" and related forms, likely indicating a focus on the term "occidental" or similar terms

No Comments

No Known Activations