INDEX

Explanations

terms related to "hat" and its variations, including forms of the word such as "hating."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

ThroughAttribute

-0.66

pherals

-0.61

 numberWith

-0.60

 portu

-0.58

rosos

-0.57

 reticulum

-0.57

️

-0.57

 Requests

-0.56

kheim

-0.56

 promis

-0.56

POSITIVE LOGITS

Hat

0.89

 hats

0.78

HAT

0.77

Hat

0.76

 brim

0.69

hat

0.68

 Hats

0.67

hat

0.58

 trick

0.58

Hats

0.57

Activations Density 0.127%

terms related to "hat" and its variations, including forms of the word such as "hating."

No Comments

No Known Activations