INDEX

Explanations

words related to various people's names, specifically "Catherine."

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly variations of the name "Catherine" and its derivatives

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

elling

-0.81

req

-0.81

nesota

-0.80

eat

-0.77

sight

-0.76

profit

-0.76

awaru

-0.73

spring

-0.73

direction

-0.71

holy

-0.70

POSITIVE LOGITS

issance

0.90

 Cort

0.83

 Catherine

0.80

 Anne

0.79

hew

0.77

 Chao

0.77

 Howell

0.77

 Parkinson

0.76

illian

0.76

 Evans

0.75

Activations Density 0.010%

words related to various people's names, specifically "Catherine."

mentions of specific names, particularly variations of the name "Catherine" and its derivatives

No Comments

No Known Activations