INDEX

Explanations

phrases related to emphasizing a particular point or giving weight to a statement

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing the importance or consequences of particular actions or situations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

LOS

-0.81

lines

-0.69

Vers

-0.68

yssey

-0.68

Adams

-0.65

MLA

-0.64

 Thames

-0.64

Vert

-0.63

Av

-0.61

Ohio

-0.61

POSITIVE LOGITS

 anymore

0.99

 whatsoever

0.94

 slightest

0.90

 anywhere

0.87

nor

0.81

theless

0.80

 anything

0.80

ifiable

0.73

 discriminating

0.73

 bother

0.72

Activations Density 2.169%

phrases related to emphasizing a particular point or giving weight to a statement

phrases emphasizing the importance or consequences of particular actions or situations

No Comments

No Known Activations