INDEX

Explanations

instances of names, specifically focusing on proper nouns and entities

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

 behav

-0.76

 disadvant

-0.71

 distingu

-0.70

 misunder

-0.70

 Ender

-0.69

 escape

-0.65

 Reply

-0.65

 independ

-0.64

AB

-0.64

 Ichigo

-0.64

POSITIVE LOGITS

EStreamFrame

0.98

milo

0.92

ForgeModLoader

0.90

Ń·

0.86

ola

0.85

 Plaza

0.81

ilogy

0.80

 TAMADRA

0.80

cci

0.79

Activations Density 0.143%

instances of names, specifically focusing on proper nouns and entities

No Comments

No Known Activations