INDEX

Explanations

references to geographical locations or specific names associated with various contexts

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

 Parenthood

-0.80

 Dying

-0.80

ãģį

-0.77

ãĥĩãĤ£

-0.75

 Millennium

-0.71

å¤©

-0.69

 DRAG

-0.68

 Nationwide

-0.66

SUP

-0.66

 Fargo

-0.65

POSITIVE LOGITS

aska

1.00

amia

1.00

bert

0.99

aji

0.97

awi

0.96

adesh

0.95

oha

0.95

ban

0.95

aya

0.94

pine

0.94

Activations Density 0.003%

references to geographical locations or specific names associated with various contexts

No Comments

No Known Activations