INDEX

Explanations

words related to stages or performances

oai_token-act-pair · gpt-3.5-turbo

references to a performance or presentation context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

anguages

-0.75

vernment

-0.71

alez

-0.69

ears

-0.67

 unequ

-0.66

ongo

-0.63

vironment

-0.62

icit

-0.61

 compuls

-0.61

umar

-0.61

POSITIVE LOGITS

 stage

1.18

stage

1.09

Stage

1.06

 Stage

0.99

 stages

0.91

 Sabha

0.78

 Frie

0.78

craft

0.75

roller

0.72

wright

0.70

Activations Density 0.012%

words related to stages or performances

references to a performance or presentation context

No Comments

No Known Activations