INDEX

Explanations

references to effort expended in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

 effort

-1.98

effort

-1.78

 Effort

-1.69

Effort

-1.47

 esfuerzo

-1.25

 esforço

-1.05

 exertion

-0.65

EFF

-0.60

 attempt

-0.60

TypedDataSet

-0.59

POSITIVE LOGITS

 nakalista

0.74

XMLSchema

0.68

WriteBarrier

0.67

 insuffisamment

0.65

+#+#

0.63

MIDDLEWARE

0.60

viewDidLoad

0.60

corrhi

0.59

شهاد

0.59

帖最后由

0.58

Activations Density 0.232%

references to effort expended in various contexts

No Comments

No Known Activations