INDEX

Explanations

phrases that indicate rankings or comparisons

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

oresc

-0.72

acked

-0.71

arant

-0.70

beh

-0.68

translation

-0.68

 SOLD

-0.67

olated

-0.67

ERC

-0.66

trial

-0.66

leeve

-0.66

POSITIVE LOGITS

ansas

0.73

 fellow

0.71

 Jasper

0.70

 Avatar

0.69

 Siri

0.68

 Coca

0.68

 Mecca

0.68

Je

0.68

 Khalid

0.67

 Stone

0.66

Activations Density 0.077%

phrases that indicate rankings or comparisons

No Comments

No Known Activations