INDEX

Explanations

words related to reduction or removal

oai_token-act-pair · gpt-3.5-turbo

the prefix "er" in various contexts and forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ership

-0.77

 Spears

-0.67

OUNT

-0.66

Premium

-0.64

kefeller

-0.63

orld

-0.63

ELS

-0.63

 Painter

-0.63

INESS

-0.62

IGHTS

-0.62

POSITIVE LOGITS

asure

1.24

asures

1.02

rant

0.98

icit

0.93

aser

0.92

ogenous

0.92

mination

0.91

asing

0.90

got

0.89

rors

0.82

Activations Density 0.016%

words related to reduction or removal

the prefix "er" in various contexts and forms

No Comments

No Known Activations