INDEX

Explanations

terms related to impact or influence

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 newUser

-0.44

 Lengkap

-0.44

/}.

-0.38

 čia

-0.37

 aikana

-0.37

ADELPHIA

-0.35

 Pakai

-0.35

)))));

-0.35

이터

-0.35

cemment

-0.35

POSITIVE LOGITS

affect

0.83

 affecting

0.80

 affects

0.77

 affect

0.76

 afect

0.73

 Affects

0.73

 Affect

0.72

 Affecting

0.72

 AFFECT

0.69

Affect

0.66

Activations Density 0.024%

terms related to impact or influence

No Comments

No Known Activations