INDEX

Explanations

attends to the token "different" from various types of tokens indicating a relationship or comparison

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-att/layer_0/width_16k/average_l0_61

Prompts (Dashboard)

16,384 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

0:0.00

1:0.01

2:0.56

3:0.01

4:0.01

5:0.00

6:0.08

7:0.05

8:0.00

9:0.00

10:0.00

11:0.00

12:0.00

13:0.07

14:0.10

15:0.00

Negative Logits

 elements

-0.47

 words

-0.47

 events

-0.46

 numbers

-0.46

 days

-0.46

 year

-0.45

 results

-0.45

 effects

-0.45

two

-0.44

 people

-0.44

POSITIVE LOGITS

CTA

0.52

their

0.52

false

0.52

findpost

0.52

IsContent

0.51

RCS

0.50

IMC

0.50

BLM

0.49

JAKARTA

0.49

Their

0.49

Activations Density 0.061%

attends to the token "different" from various types of tokens indicating a relationship or comparison

No Comments

No Known Activations