INDEX

Explanations

references to individuals in roles of communication or representation, particularly spokespersons and advisors

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

 hump

-0.73

cffff

-0.72

spir

-0.70

NetMessage

-0.67

iHUD

-0.66

dream

-0.66

 fertil

-0.65

shape

-0.64

 pity

-0.63

 gamble

-0.62

POSITIVE LOGITS

 Lauren

0.92

 Steph

0.88

 Anne

0.81

 Nicole

0.79

acy

0.78

atography

0.77

onse

0.77

 Vanessa

0.76

hips

0.76

 emailed

0.75

Activations Density 0.014%

references to individuals in roles of communication or representation, particularly spokespersons and advisors

No Comments

No Known Activations