INDEX

Explanations

references to specific years

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

sonian

-0.78

tro

-0.76

hes

-0.71

iera

-0.71

afort

-0.70

uan

-0.68

kell

-0.67

iqu

-0.66

heet

-0.65

hedral

-0.65

POSITIVE LOGITS

0.86

 onwards

0.81

0.76

 onward

0.76

 timeframe

0.74

0.70

ãĥ¼ãĥ«

0.69

0.66

Activations Density 0.007%

references to specific years

No Comments

No Known Activations