INDEX

Explanations

the word "like" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 kasarigan

-0.51

 näky

-0.45

 صوتيه

-0.45

 táctil

-0.44

eluaran

-0.42

 hänen

-0.42

 ámbitos

-0.42

CompilerServices

-0.42

linawan

-0.41

lapsingToolbar

-0.41

POSITIVE LOGITS

 unlike

0.61

unlike

0.58

 identical

0.55

similar

0.54

如同

0.52

Unlike

0.52

Similar

0.51

Same

0.50

 similar

0.50

 Unlike

0.50

Activations Density 0.094%

the word "like" in various contexts

No Comments

No Known Activations