INDEX

Explanations

attends to specific character patterns marked by the second token from non-matching tokens

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

0:0.81

1:0.00

2:0.01

3:0.01

4:0.09

5:0.02

6:0.00

7:0.02

Negative Logits

 houſe

-0.65

 purpoſe

-0.62

 ſeveral

-0.61

 pleaſure

-0.61

 Majefty

-0.60

 greateſt

-0.59

 reaſon

-0.59

 raiſ

-0.58

 ſta

-0.57

 leaſt

-0.57

POSITIVE LOGITS

oredCriteria

0.28

flo

0.27

iterranée

0.25

pa

0.25

or

0.24

posedge

0.24

0.23

get

0.23

Activations Density 8.340%

attends to specific character patterns marked by the second token from non-matching tokens

No Comments

No Known Activations