INDEX

Explanations

words related to things that are deeply connected or interrelated

oai_token-act-pair · gpt-3.5-turbo

concepts related to connection and interdependency

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

udeb

-0.92

viks

-0.79

VL

-0.79

ãĥīãĥ©

-0.73

iken

-0.64

Young

-0.64

girls

-0.64

OCK

-0.64

udo

-0.63

WAR

-0.62

POSITIVE LOGITS

 interconnected

1.10

 intertwined

1.09

 intertw

1.05

 woven

0.94

 intersect

0.92

 layers

0.91

 seamlessly

0.88

 intersections

0.87

 strands

0.86

 synerg

0.85

Activations Density 0.048%

words related to things that are deeply connected or interrelated

concepts related to connection and interdependency

No Comments

No Known Activations