INDEX

Explanations

phrases that indicate categories or types

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ProtoMessage

-0.48

遺

-0.48

waltung

-0.48

เท่า

-0.47

Dabei

-0.47

"""

-0.47

"",

-0.46

Xna

-0.45

 alguns

-0.45

 consecu

-0.44

POSITIVE LOGITS

########.

0.86

sizeCache

0.85

somewhere

0.81

 Something

0.80

Something

0.79

ladesh

0.76

 Somewhere

0.76

 wireType

0.74

Somebody

0.74

SpringBootTest

0.74

Activations Density 0.151%

phrases that indicate categories or types

No Comments

No Known Activations