INDEX

Explanations

phrases related to manipulation and control, particularly in a narrative context

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

tool or pawn

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ianto

-0.35

 bekijken

-0.35

łoży

-0.35

 wyposa

-0.34

berapa

-0.33

 تحد

-0.33

 acciaio

-0.33

RET

-0.33

án

-0.33

read

-0.32

POSITIVE LOGITS

 cherchés

0.71

ValueStyle

0.57

 Taktlose

0.56

 للاسماء

0.54

 ObjectParameter

0.53

TypedDataSet

0.52

__":

0.49

jsonwebtoken

0.49

ỡng

0.48

AsUp

0.48

Activations Density 0.059%

phrases related to manipulation and control, particularly in a narrative context

tool or pawn

No Comments

No Known Activations