INDEX

Explanations

actions and claims related to teachings, beliefs, and interpretations of authority figures or texts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

SHARE

-0.77

VIDEO

-0.76

BuyableInstoreAndOnline

-0.75

worth

-0.75

xon

-0.73

cv

-0.70

rir

-0.68

pee

-0.68

ICO

-0.68

FACE

-0.66

POSITIVE LOGITS

 Parables

0.81

 analogous

0.80

 philosophers

0.79

 explicitly

0.77

 doctrines

0.76

 notions

0.75

 textual

0.75

 correlations

0.71

 distinctions

0.71

 adherents

0.71

Activations Density 0.180%

actions and claims related to teachings, beliefs, and interpretations of authority figures or texts

No Comments

No Known Activations