INDEX

Explanations

various instances of the word "ideas"

oai_token-act-pair · gpt-3.5-turbo

references to innovative or philosophical concepts

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

ded

-0.71

Delivery

-0.70

Military

-0.67

 Mamm

-0.66

ORD

-0.65

 occupancy

-0.64

ALL

-0.63

sole

-0.61

enary

-0.60

effect

-0.60

POSITIVE LOGITS

ensical

0.92

 ideas

0.84

matter

0.83

ĸļ

0.81

uggest

0.81

mith

0.78

omething

0.77

 Ideas

0.75

cape

0.74

 theories

0.74

Activations Density 0.064%

various instances of the word "ideas"

references to innovative or philosophical concepts

No Comments

No Known Activations