INDEX

Explanations

phrases related to objectives or goals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 plomo

-0.40

 öğretmen

-0.36

 coordinación

-0.35

 urusan

-0.34

 negocia

-0.34

 hatır

-0.34

 sürd

-0.32

 llorar

-0.32

meyen

-0.32

 düşünü

-0.32

POSITIVE LOGITS

TagMode

0.74

 nodeId

0.69

 nodes

0.63

 density

0.63

 node

0.60

 Manuscript

0.60

Scalar

0.60

 Scalar

0.60

Node

0.59

 goal

0.59

Activations Density 0.232%

phrases related to objectives or goals

No Comments

No Known Activations