INDEX

Explanations

occurrences of the letter 'e' and related derivatives in various contexts

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

preceding "closed" or "jections"

np_acts-logits-general · gemini-2.0-flash

ecdysone, eclosion, ejections

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 myſelf

-0.57

terase

-0.50

мулятор

-0.49

Efq

-0.49

msgTypes

-0.48

 Majefty

-0.48

 itſelf

-0.48

 oignon

-0.47

 beſ

-0.47

下载附件

-0.47

POSITIVE LOGITS

ManagedBean

0.59

utek

0.58

Rüyada

0.56

AutoScaleMode

0.53

Chop

0.51

UnknownFieldSet

0.51

makeText

0.51

 Walkover

0.50

 متعلقه

0.50

BeginContext

0.50

Activations Density 0.329%

occurrences of the letter 'e' and related derivatives in various contexts

preceding "closed" or "jections"

ecdysone, eclosion, ejections

No Comments

No Known Activations