INDEX

Explanations

phrases related to binding or being bound

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of "binding."

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

cano

-0.78

ODUCT

-0.75

asury

-0.74

akura

-0.70

amera

-0.69

ortmund

-0.68

zanne

-0.66

issance

-0.65

ç¥ŀ

-0.64

PRO

-0.64

POSITIVE LOGITS

sym

1.18

 bound

1.08

ingly

1.00

 bind

0.99

binding

0.95

 binds

0.93

masters

0.81

fold

0.79

 binding

0.78

nect

0.78

Activations Density 0.008%

phrases related to binding or being bound

terms related to the concept of "binding."

No Comments

No Known Activations