INDEX

Explanations

phrases or expressions indicating availability or suitability

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

mare

-0.68

çīĪ

-0.65

nik

-0.64

 Originally

-0.64

 Ahead

-0.63

 beware

-0.62

 GOODMAN

-0.62

isky

-0.61

natureconservancy

-0.60

ouf

-0.59

POSITIVE LOGITS

 sale

1.07

cible

0.93

 grabs

0.92

geries

0.92

 inspection

0.88

ked

0.88

 consideration

0.88

 reuse

0.88

 purchase

0.81

 viewing

0.80

Activations Density 0.112%

phrases or expressions indicating availability or suitability

No Comments

No Known Activations