INDEX

Explanations

occurrences of numerical references and counts in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ͡°)

-0.64

pannt

-0.60

SequentialGroup

-0.56

nicity

-0.54

llary

-0.53

")))

-0.53

calypse

-0.52

]--;

-0.52

]++;

-0.51

vajal

-0.51

POSITIVE LOGITS

ViewImports

0.84

AndEndTag

0.74

 countless

0.68

几次

0.65

InputBorder

0.63

StoryboardSegue

0.62

 numerous

0.59

NKC

0.56

Numerous

0.56

+#+

0.55

Activations Density 0.291%

occurrences of numerical references and counts in various contexts

No Comments

No Known Activations