INDEX

Explanations

conjunctions indicating a conclusion or inference

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

introducing conclusions or consequences

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

 blin

-0.64

Arti

-0.58

ScopeManager

-0.57

 parlor

-0.57

 trico

-0.56

PSF

-0.55

BrowserModule

-0.55

ClickListener

-0.55

VCR

-0.55

Pw

-0.54

POSITIVE LOGITS

 Hence

1.67

Hence

1.66

 hence

1.58

hence

1.49

 pertanto

0.76

 perciò

0.71

 appunto

0.69

 therefore

0.68

 henceforth

0.68

whence

0.67

Activations Density 0.022%

conjunctions indicating a conclusion or inference

introducing conclusions or consequences

No Comments

No Known Activations