INDEX

Explanations

phrases related to personal names

oai_token-act-pair · gpt-3.5-turbo

frequent names and terms related to specific individuals and their actions

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

DragonMagazine

-0.69

ADRA

-0.65

alore

-0.63

achus

-0.62

CFR

-0.62

ascus

-0.61

abase

-0.61

ptroller

-0.60

 ILCS

-0.59

 indo

-0.59

POSITIVE LOGITS

ħĭ

0.84

velt

0.70

 heights

0.69

Ħ¢

0.66

*=-

0.64

pal

0.63

aman

0.61

 Ital

0.61

hani

0.60

¿

0.60

Activations Density 0.198%

phrases related to personal names

frequent names and terms related to specific individuals and their actions

No Comments

No Known Activations