INDEX

Explanations

phrases related to legal entities or agreements

oai_token-act-pair · gpt-3.5-turbo

instances of a specific abbreviation or term

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

mable

-0.83

croft

-0.82

manship

-0.81

matically

-0.80

shire

-0.79

lime

-0.79

pour

-0.77

meyer

-0.77

south

-0.74

case

-0.73

POSITIVE LOGITS

ILS

0.91

IL

0.83

chnology

0.77

ival

0.77

ired

0.71

IRED

0.69

roma

0.69

 instincts

0.67

 Polo

0.65

uder

0.65

Activations Density 0.026%

phrases related to legal entities or agreements

instances of a specific abbreviation or term

No Comments

No Known Activations