INDEX

Explanations

names related to specific individuals

oai_token-act-pair · gpt-3.5-turbo

references to specific names and terms related to individuals and organizations

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

OW

-0.79

oris

-0.75

prints

-0.70

inctions

-0.68

oro

-0.67

rial

-0.65

 ticket

-0.65

iago

-0.64

otal

-0.63

Ú

-0.63

POSITIVE LOGITS

ldom

0.81

beck

0.79

clinton

0.79

nesday

0.76

tarians

0.74

osponsors

0.73

Zel

0.73

marine

0.72

ben

0.72

earchers

0.72

Activations Density 0.044%

names related to specific individuals

references to specific names and terms related to individuals and organizations

No Comments

No Known Activations