INDEX

Explanations

the occurrences of the pronoun "I."

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

é¾įå¥ĳå£«

-0.65

rules

-0.63

 waivers

-0.61

 neutrality

-0.59

 transports

-0.58

 reimbursement

-0.57

 apex

-0.56

 dimensions

-0.56

 extant

-0.56

 forks

-0.55

POSITIVE LOGITS

stanbul

1.14

ATA

1.13

RA

1.13

EEE

1.11

'm

1.09

BA

1.07

FL

1.03

nex

1.02

KE

1.01

OV

1.01

Activations Density 0.038%

the occurrences of the pronoun "I."

No Comments

No Known Activations