INDEX

Explanations

intentions and goals expressed through the use of the word "to"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

󠁴

-0.71

 nakalista

-0.70

FromNib

-0.65

-0.62

parseLong

-0.60

thyl

-0.58

bobby

-0.58

 Rumble

-0.57

 jamb

-0.56

 Squirrel

-0.56

POSITIVE LOGITS

 aims

0.63

是为了

0.56

 visant

0.54

 nhằm

0.51

 ensuring

0.51

 ensure

0.51

 aimed

0.50

 objectif

0.48

目的是

0.46

aim

0.45

Activations Density 0.028%

intentions and goals expressed through the use of the word "to"

No Comments

No Known Activations