INDEX

Explanations

phrases that mention numerical values or comparative relationships

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

stocks

-0.80

ensor

-0.79

ENS

-0.77

ens

-0.70

={

-0.68

{{

-0.67

avid

-0.65

Spoiler

-0.65

rouse

-0.65

hei

-0.63

POSITIVE LOGITS

 others

1.00

 Others

0.89

 another

0.87

 secondly

0.87

 Another

0.85

Others

0.73

one

0.70

one

0.69

 remainder

0.67

 Secondly

0.67

Activations Density 0.204%

phrases that mention numerical values or comparative relationships

No Comments

No Known Activations