INDEX

Explanations

phrases expressing continuity and wholeness in experiences or narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Direkt

-0.56

Cog

-0.52

ilar

-0.50

hdl

-0.49

opat

-0.48

lar

-0.48

icide

-0.48

ops

-0.48

pa

-0.47

TEntity

-0.47

POSITIVE LOGITS

throughout

1.34

 throughout

1.26

 Throughout

1.11

Throughout

1.07

 boyunca

1.05

 sepanjang

1.01

 suốt

0.94

 hinweg

0.83

 THRO

0.82

 entire

0.81

Activations Density 0.237%

phrases expressing continuity and wholeness in experiences or narratives

No Comments

No Known Activations