INDEX

Explanations

words that indicate uncertainty or subjectivity

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

adelphia

-0.83

anamo

-0.77

 Horizons

-0.77

waukee

-0.75

iets

-0.74

iens

-0.74

xia

-0.74

uay

-0.74

ouses

-0.73

swick

-0.72

POSITIVE LOGITS

 alike

0.82

 profitable

0.79

 contradictory

0.75

 camoufl

0.74

 consequential

0.73

 downright

0.72

-"

0.72

 depend

0.71

 fatal

0.71

)-

0.71

Activations Density 0.097%

words that indicate uncertainty or subjectivity

No Comments

No Known Activations