INDEX

Explanations

expressions related to progression or directionality in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.17.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

 itſelf

-1.10

Efq

-0.96

 themſelves

-0.93

ſelves

-0.89

 Shakspeare

-0.89

 myſelf

-0.88

 Anſ

-0.87

 raiſ

-0.86

 againſt

-0.85

ValueStyle

-0.84

POSITIVE LOGITS

the

1.36

be

1.07

 this

0.97

 make

0.94

an

0.86

0.85

 give

0.83

our

0.78

 take

0.77

 bring

0.77

Activations Density 0.305%

expressions related to progression or directionality in context

No Comments

No Known Activations