INDEX

Explanations

items or instructions marked as "optional."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "optional" and its variations

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

imaru

-0.81

masters

-0.79

builders

-0.78

ells

-0.78

Building

-0.77

BW

-0.77

uay

-0.77

abal

-0.77

lins

-0.76

wordpress

-0.75

POSITIVE LOGITS

 extras

0.97

 extra

0.89

 accessory

0.85

 optional

0.83

 eaves

0.82

 supplemental

0.77

 accessories

0.75

 intermediate

0.74

 bonus

0.74

 foil

0.73

Activations Density 0.028%

items or instructions marked as "optional."

instances of the word "optional" and its variations

No Comments

No Known Activations