INDEX

Explanations

phrases related to drawing or capturing attention

oai_token-act-pair · gpt-3.5-turbo

mentions of attention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Mehran

-0.70

tre

-0.70

 halves

-0.68

iche

-0.64

imar

-0.62

 Penal

-0.62

tein

-0.61

 Staten

-0.60

Lie

-0.59

UES

-0.59

POSITIVE LOGITS

 attention

1.06

 Attention

0.92

ively

0.87

estinal

0.87

 scrutiny

0.81

 attentive

0.78

orial

0.77

ibility

0.75

ipal

0.71

ense

0.71

Activations Density 0.021%

phrases related to drawing or capturing attention

mentions of attention

No Comments

No Known Activations