INDEX

Explanations

expressions or vocabulary associated with positive experiences or feedback

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sur

-0.46

jura

-0.45

舌

-0.43

〔

-0.42

chok

-0.42

bac

-0.41

rage

-0.41

Tcl

-0.41

tencent

-0.41

tam

-0.41

POSITIVE LOGITS

principalTable

0.75

MergeFrom

0.69

 kasarigan

0.67

modelBuilder

0.67

 mergeFrom

0.66

 Extragalactic

0.65

 contextLoads

0.64

PerformLayout

0.59

 asmen

0.58

featureID

0.58

Activations Density 0.276%

expressions or vocabulary associated with positive experiences or feedback

No Comments

No Known Activations