INDEX

Explanations

prepositions with a reference to spatial positions indicating the location of an object relative to the speaker

oai_token-act-pair · gpt-3.5-turbo

pronouns denoting people in various 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

 Compliance

-0.61

 Majority

-0.60

 Serious

-0.59

uncture

-0.59

 fact

-0.58

UFF

-0.58

 majority

-0.57

 Heist

-0.56

AAA

-0.56

 Podesta

-0.55

POSITIVE LOGITS

selves

1.15

self

1.08

 selves

0.91

atic

0.88

atically

0.88

 personally

0.81

andering

0.73

alian

0.73

 orally

0.72

entric

0.71

Activations Density 0.112%

prepositions with a reference to spatial positions indicating the location of an object relative to the speaker

pronouns denoting people in various contexts

No Comments

No Known Activations