INDEX

Explanations

names of people or organizations

oai_token-act-pair · gpt-3.5-turbo

phrases that express comparisons or likening to other entities

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

ascript

-0.78

iven

-0.71

ysical

-0.71

Closure

-0.69

 reversible

-0.67

VERTISEMENT

-0.65

Characters

-0.63

ells

-0.63

ipple

-0.63

 unsolved

-0.63

POSITIVE LOGITS

lier

0.92

Cec

0.83

 Jacob

0.78

 Mish

0.77

Er

0.77

 Pamela

0.77

Ian

0.76

 Mike

0.76

Muk

0.75

lihood

0.74

Activations Density 0.109%

names of people or organizations

phrases that express comparisons or likening to other entities

No Comments

No Known Activations