INDEX

Explanations

references to "high" in various contexts, particularly regarding education and performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Administrativna

-0.74

 '\\;'

-0.70

ftagPool

-0.67

 MainAxisSize

-0.64

 extranjera

-0.63

oltre

-0.62

RTDA

-0.61

 étrangère

-0.59

Hochspringen

-0.59

 podjela

-0.58

POSITIVE LOGITS

 High

0.87

High

0.84

 school

0.82

high

0.74

school

0.74

 high

0.73

lighting

0.70

lighted

0.70

 School

0.70

 HIGH

0.68

Activations Density 0.165%

references to "high" in various contexts, particularly regarding education and performance

No Comments

No Known Activations