INDEX

Explanations

phrases that indicate percentages or statistics related to activities or events

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

 advoc

-0.72

 incorpor

-0.72

raq

-0.70

 offending

-0.69

 enlarg

-0.68

 appro

-0.67

 indu

-0.67

 incre

-0.66

 independ

-0.65

ppe

-0.65

POSITIVE LOGITS

 Latest

0.84

 Thousands

0.83

 Authorities

0.82

 Officials

0.81

 Protesters

0.79

 Despite

0.79

 Recap

0.74

Imagine

0.74

 Saying

0.74

[[

0.74

Activations Density 0.010%

phrases that indicate percentages or statistics related to activities or events

No Comments

No Known Activations