INDEX

Explanations

directional terms or references in a text

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

ulous

-0.74

inous

-0.73

iasis

-0.64

 bearer

-0.61

 constants

-0.59

urion

-0.59

inance

-0.58

alien

-0.57

aceae

-0.57

chio

-0.56

POSITIVE LOGITS

ward

1.34

wards

1.22

west

1.05

outheast

1.01

side

1.01

bound

0.93

WARD

0.87

forth

0.87

etheless

0.86

 west

0.84

Activations Density 0.014%

directional terms or references in a text

No Comments

No Known Activations