INDEX

Explanations

references to walls and wall-related concepts

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

TagMode

-0.58

 propOrder

-0.54

 nahilalakip

-0.47



-0.47

 utafitiHapana

-0.47

TargetException

-0.47

hoeddwyd

-0.46

 насељу

-0.44

TypedDataSet

-0.43

CppMethod

-0.43

POSITIVE LOGITS

 wall

0.67

 hang

0.58

 murale

0.55

 Hang

0.51

 Wall

0.50

 WALL

0.50

 hanging

0.49

 walls

0.49

 AttributeSet

0.49

Wall

0.47

Activations Density 0.005%

references to walls and wall-related concepts

No Comments

No Known Activations