INDEX

Explanations

phrases related to long-term outcomes or decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_4/width_65k/average_l0_66

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.4.hook_mlp_out

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

 Spuren

-0.71

 Signalez

-0.65

CloseOperation

-0.62

saraba

-0.61

 signs

-0.59

########.

-0.59

 traces

-0.58

guenos

-0.56

ijnen

-0.55

참고

-0.55

POSITIVE LOGITS

DockStyle

0.79

 ultimately

0.69

StoryboardSegue

0.66

Ultimately

0.65

 Ultimately

0.64

 ulti

0.57

LabelTagHelper

0.54

 uiteindelijk

0.54

 Ontology

0.53

ambilan

0.53

Activations Density 0.066%

phrases related to long-term outcomes or decisions

No Comments

No Known Activations