INDEX

Explanations

phrases indicating discomfort or distress

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 genannt

-0.54

hiva

-0.53

 purpure

-0.53

balah

-0.52

felizmente

-0.51

WRENCE

-0.50

 nito

-0.50

wla

-0.50

 péché

-0.50

hamshire

-0.47

POSITIVE LOGITS

 Numerade

0.67

 للاسماء

0.67

fjspx

0.61

".

0.59

 unknownFields

0.58

SourceChecksum

0.58

 sensed

0.57

 noticing

0.56

 suspicious

0.56

ніципа

0.56

Activations Density 0.335%

phrases indicating discomfort or distress

No Comments

No Known Activations