INDEX

Explanations

contrasts and comparisons in language

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

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

Negative Logits

nodoc

-0.69

SequentialGroup

-0.69

mybatisplus

-0.65

<pad>

-0.62

<unused16>

-0.61

<unused14>

-0.61

<unused23>

-0.61

<unused28>

-0.61

[@BOS@]

-0.61

<unused8>

-0.61

POSITIVE LOGITS

而

0.44

 whereas

0.44

Whereas

0.39

TED

0.36

 Whereas

0.35

而

0.34

LO

0.33

 others

0.33

 invece

0.33

whereas

0.32

Activations Density 0.832%

contrasts and comparisons in language

No Comments

No Known Activations