INDEX

Explanations

phrases containing the specific word "morrow."

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of "moron" or derogatory descriptions of people

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

BOOK

-0.79

XY

-0.67

ECH

-0.66

 Wheeler

-0.65

sets

-0.65

 Carbuncle

-0.64

orney

-0.63

Brand

-0.61

Book

-0.61

 Piercing

-0.60

POSITIVE LOGITS

gue

1.09

aceutical

0.98

gage

0.92

iday

0.92

itary

0.87

ose

0.84

idian

0.81

 surv

0.80

agi

0.79

mor

0.79

Activations Density 0.016%

phrases containing the specific word "morrow."

terms related to the concept of "moron" or derogatory descriptions of people

No Comments

No Known Activations