INDEX

Explanations

expressions of vulnerability and openness in communication

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

 informée

-0.71

 surla

-0.69

 MainAxisSize

-0.64

 ujednoznacz

-0.63

utives

-0.62

balleur

-0.61

ikusbot

-0.59

 للمعارف

-0.58

iffance

-0.58

 المعيارى

-0.57

POSITIVE LOGITS

 shared

0.42

 vulnerability

0.40

 revealing

0.40

 sharing

0.39

 compartió

0.38

shared

0.35

Sharing

0.34

 intimate

0.34

 condiv

0.34

 openness

0.33

Activations Density 0.019%

expressions of vulnerability and openness in communication

No Comments

No Known Activations