INDEX

Explanations

past tense verbs and phrases indicating ongoing or past actions

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

strconv

-0.61

OPINION

-0.57

ähig

-0.56

softc

-0.56

wpi

-0.56

Grüsse

-0.53

berdayakan

-0.53

 désormais

-0.52

]>=

-0.52

pozdrawiam

-0.51

POSITIVE LOGITS

 thinking

0.75

 referenties

0.72

 reading

0.67

 hoping

0.65

 told

0.64

InjectAttribute

0.61

hop

0.60

 lenker

0.60

thinking

0.58

 végét

0.58

Activations Density 0.096%

past tense verbs and phrases indicating ongoing or past actions

No Comments

No Known Activations