INDEX

Explanations

references to "closed" scenarios or situations

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

iage

-0.85

nesota

-0.74

intent

-0.70

aire

-0.70

andum

-0.70

ãĥīãĥ©ãĤ´ãĥ³

-0.69

ãĤ¼ãĤ¦ãĤ¹

-0.69

CTV

-0.68

 Metatron

-0.68

omething

-0.65

POSITIVE LOGITS

 doors

0.93

 minded

0.91

 door

0.85

 loopholes

0.81

lid

0.80

 Caption

0.79

 eyel

0.77

 Doors

0.75

 caption

0.74

 pores

0.72

Activations Density 0.011%

references to "closed" scenarios or situations

No Comments

No Known Activations