INDEX

Explanations

references to dependency and conditions influencing outcomes

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

BuyableInstoreAndOnline

-0.89

ãĥĨ

-0.89

ãĥĥãĥī

-0.87

ãĤ§

-0.86

ãĥ¼ãĥ³

-0.86

MpServer

-0.86

ãĥ´

-0.82

Cola

-0.82

audi

-0.82

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.81

POSITIVE LOGITS

 whether

0.98

 persu

0.95

 maintaining

0.94

 factors

0.94

 figuring

0.91

 ensuring

0.89

 convincing

0.88

 respecting

0.88

 patience

0.87

 obtaining

0.86

Activations Density 0.028%

references to dependency and conditions influencing outcomes

No Comments

No Known Activations