INDEX

Explanations

phrases or words that are specified or indicated in a detailed manner

oai_token-act-pair · gpt-3.5-turbo

instances of the word "specifically."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lyn

-0.73

 Tycoon

-0.73

ulton

-0.69

 Isles

-0.68

anon

-0.66

IAN

-0.66

 Neighbor

-0.65

 Kenn

-0.65

ocene

-0.64

 Hands

-0.64

POSITIVE LOGITS

 tailored

1.08

cedented

0.95

 identifiable

0.92

 targeted

0.91

 exempted

0.89

 tuned

0.87

 suited

0.85

 guiActiveUn

0.84

 formulated

0.82

 geared

0.81

Activations Density 0.010%

phrases or words that are specified or indicated in a detailed manner

instances of the word "specifically."

No Comments

No Known Activations