INDEX

Explanations

instances of the words "be" and "to."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

밌

-0.46

 lots

-0.45

 back

-0.44

 Ironically

-0.43

 geri

-0.43

 occhiali

-0.42

dorfer

-0.42

ț

-0.41

 dependent

-0.41

 Depending

-0.41

POSITIVE LOGITS

GEBURTSDATUM

1.12

 AssemblyProduct

1.03

 AssemblyTitle

1.02

UrlResolution

0.98

 Chwiliwch

0.96

 unknownFields

0.96

<bos>

0.95

uxxxx

0.94

TypedDataSet

0.93

InitVars

0.93

Activations Density 0.276%

instances of the words "be" and "to."

No Comments

No Known Activations