INDEX

Explanations

phrases indicating actions directed towards a target, particularly in a combat or conflict context

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

at followed by articles or pronouns

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

principalColumn

-0.45

Personendaten

-0.42

 erschienen

-0.40

HasBeenSet

-0.40

 honored

-0.38

География

-0.37

 kasarigan

-0.37

 esternos

-0.37

dư

-0.37

UserScript

-0.36

POSITIVE LOGITS

Towards

0.69

 Towards

0.67

 towards

0.63

towards

0.62

 Toward

0.61

 toward

0.57

Toward

0.55

 staring

0.55

findpost

0.53

toward

0.51

Activations Density 0.045%

phrases indicating actions directed towards a target, particularly in a combat or conflict context

at followed by articles or pronouns

No Comments

No Known Activations