INDEX

Explanations

phrases related to being well-informed

oai_token-act-pair · gpt-3.5-turbo

the term "informed" in various contexts related to knowledge and awareness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ingu

-0.88

enf

-0.69

imposed

-0.66

cessive

-0.66

Untitled

-0.64

EG

-0.62

pen

-0.61

cdn

-0.61

Nice

-0.60

Static

-0.60

POSITIVE LOGITS

 informed

0.87

 inform

0.86

ally

0.84

 Consent

0.81

ingly

0.80

 assurance

0.79

omic

0.79

soType

0.76

fulness

0.76

 incorpor

0.73

Activations Density 0.011%

phrases related to being well-informed

the term "informed" in various contexts related to knowledge and awareness

No Comments

No Known Activations