INDEX

Explanations

instances of the word "follow" in various 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

astă

-0.62

seende

-0.55

OptionsMenu

-0.53

 BoxFit

-0.52

enment

-0.52

cadora

-0.51

än

-0.51

beginn

-0.51

 bahawa

-0.51

_^

-0.50

POSITIVE LOGITS

 suit

1.26

 Follows

1.03

 closely

0.87

 follows

0.87

GraphicsUnit

0.84

Followed

0.84

suit

0.82

follows

0.81

 FOLLOW

0.81

follow

0.81

Activations Density 0.126%

instances of the word "follow" in various contexts

No Comments

No Known Activations