INDEX

Explanations

terms related to magnification or emphasis

oai_token-act-pair · gpt-3.5-turbo

terms related to "magnitude" or "magnification."

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

OPLE

-0.92

BOOK

-0.90

 Alive

-0.82

pta

-0.78

 Warriors

-0.72

IRO

-0.70

FAULT

-0.69

 Samurai

-0.69

FO

-0.69

ndra

-0.68

POSITIVE LOGITS

olia

1.15

ific

1.13

ificent

1.13

itudes

1.09

animous

1.02

amic

0.98

ately

0.93

 glim

0.92

aques

0.89

itud

0.89

Activations Density 0.018%

terms related to magnification or emphasis

terms related to "magnitude" or "magnification."

No Comments

No Known Activations