INDEX

Explanations

references to watches, wearables, and their features or functionalities

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

Reviewer

-0.99

«ĺ

-0.85

ajor

-0.80

EngineDebug

-0.79

NetMessage

-0.71

staking

-0.71

ãĥ£

-0.71

 defe

-0.69

REDACTED

-0.68

innie

-0.66

POSITIVE LOGITS

tower

1.07

maker

1.04

dogs

1.01

dog

0.97

cases

0.88

faces

0.88

marks

0.85

makers

0.85

watch

0.81

pad

0.80

Activations Density 0.012%

references to watches, wearables, and their features or functionalities

No Comments

No Known Activations