INDEX

Explanations

countries and political figures mentioned in a document

oai_token-act-pair · gpt-3.5-turbo

names and references to individuals in political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ensibly

-0.79

pires

-0.71

encia

-0.66

Sphere

-0.66

ttle

-0.65

mire

-0.64

soever

-0.62

until

-0.62

pure

-0.61

mediate

-0.61

POSITIVE LOGITS

 proverb

0.99

 remark

0.94

 saying

0.93

 precedent

0.89

 noting

0.89

 similarities

0.88

 example

0.86

 accomplishments

0.86

 preced

0.86

 stating

0.85

Activations Density 0.434%

countries and political figures mentioned in a document

names and references to individuals in political contexts

No Comments

No Known Activations