INDEX

Explanations

words indicative of theoretical discussions or reviews in academic contexts

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

another

-0.53

 another

-0.52

 weiteren

-0.52

 weitere

-0.51

 Another

-0.51

Further

-0.50

 ytterligare

-0.50

 autre

-0.50

 weiterer

-0.50

還能

-0.49

POSITIVE LOGITS

 starts

2.39

 start

2.29

 starting

2.19

 begin

2.08

 begins

2.05

 Start

2.05

 Starting

2.04

 Starts

2.03

 started

2.01

 dimulai

1.99

Activations Density 0.469%

words indicative of theoretical discussions or reviews in academic contexts

No Comments

No Known Activations