INDEX

Explanations

attends to the token "linear" found in various contexts from the punctuation mark "period."

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.attn.hook_z

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

Head Attr Weights

0:0.01

1:0.02

2:0.04

3:0.66

4:0.04

5:0.03

6:0.11

7:0.05

Negative Logits

 myſelf

-1.27

Efq

-1.26

 itſelf

-1.23

 purpoſe

-1.20

 Monfieur

-1.18

 pleaſure

-1.16

 themſelves

-1.12

 ſeveral

-1.11

 Majefty

-1.11

 Jefus

-1.11

POSITIVE LOGITS

0.75

0.60

0.59

0.57

0.54

0.53

0.52

«

0.52

Activations Density 1.571%

attends to the token "linear" found in various contexts from the punctuation mark "period."

No Comments

No Known Activations